{"forums":[{"id":4024,"title":"Visual Intelligence Meets Multimodality: From Perception to Understanding","topic":"Vision, Multimodality and World Models","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-108邛海","description":"With the explosive growth of generative AI and multimodal large models, visual intelligence and multimodal learning have become an important research frontier and industrial breakthrough in the field of artificial intelligence. Vision is not only the foundation of human perception, but also the core of multimodal understanding with the highest complexity and the greatest room for innovation. Currently, the integration of vision with text, audio, and physical behavior is accelerating, leading AI to achieve a paradigm shift from \"passive perception\" to \"active understanding and collaborative decision-making.\" This Tutorial focuses on \"Visual Intelligence Meets Multimodality: From Perception to Understanding,\" systematically analyzing key technologies such as cross-modal generalization, multimodal collaboration, collaborative generation, and embodied interaction, and, combined with cutting-edge research, building a complete technical map for participants. The academic exchange will pay special attention to cross-modal robust alignment, high-dimensional feature fusion, and multi-task linkage, aiming to solve the generalization bottleneck in open environments and jointly promote the industrial implementation of the next-generation multimodal intelligence ecosystem.","is_demo":0,"cover_url":"/api/forums/4024/cover?size=thumb","cover_alt":"Visual Intelligence Meets Multimodality: From Perception to Understanding","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4024","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4024","forum_type":"专题论坛","format":"Tutorial","theme_code":"视觉、多模态与世界模型","theme_name":"视觉、多模态与世界模型","venue":"天府国际会议中心","room":"天府国际会议中心-108邛海","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4024","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"视觉智能邂逅多模态：从感知到理解","original_topic":"视觉、多模态与世界模型","original_description":"伴随生成式AI与多模态大模型的爆发式增长，视觉智能与多模态学习已成为人工智能领域的重要研究前沿和产业突破口。视觉不仅是人类感知的基础，更是多模态理解中复杂度最高、创新空间最大的核心。当前，视觉与文本、音频、物理行为的融合正加速演进，引领AI实现从“被动感知”到“主动理解与协同决策”的范式转变。本次Tutorial聚焦“视觉智能邂逅多模态：从感知到理解”，系统剖析跨模态泛化、多模态协同、协同生成及具身交互等关键技术，结合前沿研究，为参会者构建完整的技术图谱。学术交流将特别关注跨模态鲁棒对齐、高维特征融合与多任务联动，旨在解决开放环境下的泛化瓶颈，共同推动下一代多模态智能生态的产业落地。","people":[{"forum_id":4024,"role":"guest","name":"张长青","affiliation":"天津大学","sort_order":0},{"forum_id":4024,"role":"guest","name":"马佳义","affiliation":"武汉大学","sort_order":1},{"forum_id":4024,"role":"guest","name":"左旺孟","affiliation":"哈尔滨工业大学","sort_order":2},{"forum_id":4024,"role":"guest","name":"江俊君","affiliation":"哈尔滨工业大学","sort_order":3},{"forum_id":4024,"role":"guest","name":"江奎","affiliation":"哈尔滨工业大学","sort_order":4},{"forum_id":4024,"role":"guest","name":"贾修一","affiliation":"南京理工大学","sort_order":5}],"host":"","image_url":"/api/forums/4024/cover"},{"id":4028,"title":"Dynamic Graph Representation Learning: Structure, Interaction, Semantics, and Applications","topic":"Unclassified","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-112光雾山","description":"Real-world relational data typically has continuously evolving structures, temporal interactions, and multiple types of semantics, driving graph learning from static structure modeling toward dynamic process representation. Centered on the core question of \"how to uniformly represent temporal evolution and heterogeneous semantics,\" this tutorial systematically introduces the main modeling paradigms of discrete-time dynamic graphs, continuous-time dynamic graphs, and dynamic heterogeneous graphs, connecting the three technical main lines of structural evolution, temporal interaction, and semantic modeling. The tutorial further combines high-impact applications such as financial risk and anomaly detection to discuss key issues such as benchmarking, scalability, generalization, and interpretability, and, oriented toward large models and graph foundation models, explores the development direction of dynamic graph representation toward general dynamic cognitive modeling.","is_demo":0,"cover_url":"/api/forums/4028/cover?size=thumb","cover_alt":"Dynamic Graph Representation Learning: Structure, Interaction, Semantics, and Applications","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4028","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4028","forum_type":"专题论坛","format":"Tutorial","theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-112光雾山","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4028","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"动态图表征学习：结构、交互、语义与应用","original_topic":"未分类","original_description":"现实世界中的关联数据通常具有持续演化的结构、时序交互与多类型语义，推动图学习从静态结构建模走向动态过程表征。本教程围绕“如何统一表征时间演化与异质语义”这一核心问题，系统介绍离散时间动态图、连续时间动态图与动态异质图的主要建模范式，贯通结构演化、时序交互与语义建模三条技术主线。教程进一步结合金融风险、异常检测等高影响应用，讨论基准评测、可扩展性、泛化性、可解释性等关键问题，并面向大模型与图基础模型，探讨动态图表征迈向通用动态认知建模的发展方向。","people":[{"forum_id":4028,"role":"guest","name":"焦鹏飞","affiliation":"杭州电子科技大学","sort_order":0},{"forum_id":4028,"role":"guest","name":"袁增辉","affiliation":"杭州电子科技大学","sort_order":1},{"forum_id":4028,"role":"guest","name":"陈可倪","affiliation":"杭州电子科技大学","sort_order":2},{"forum_id":4028,"role":"guest","name":"郑蕾蕾","affiliation":"杭州电子科技大学","sort_order":3},{"forum_id":4028,"role":"guest","name":"刘欢","affiliation":"杭州电子科技大学","sort_order":4},{"forum_id":4028,"role":"guest","name":"陈竑江","affiliation":"杭州电子科技大学","sort_order":5},{"forum_id":4028,"role":"guest","name":"金弟","affiliation":"天津大学","sort_order":6}],"host":"","image_url":"/api/forums/4028/cover"},{"id":4030,"title":"Probabilistic Trustworthy Learning in the Era of Large Models: Uncertainty, Calibration, and Reliable Reasoning","topic":"Unclassified","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-107A三星堆A","description":"With the rapid development of large models, multimodal large models, and agent systems, AI is moving from \"answering questions\" to \"expressing reasoning, invoking tools, and participating in decision-making.\" But a prominent problem is: large models often do not know when they do not know—they may answer confidently on unknown questions, generate seemingly plausible but incorrect chains of thought in complex reasoning, be disturbed by evidence conflicts and visual hallucinations in multimodal scenarios, and lack necessary mechanisms for verification, refusal to answer, and risk perception. The key to trustworthy large models is no longer just improving capability scale, but enabling models to recognize the unknown, express uncertainty, calibrate confidence, and adopt safer reasoning and decision-making strategies when unreliable. Probabilistic learning has long focused on uncertainty modeling, confidence calibration, and reliable decision-making, providing systematic theoretical and methodological tools for this. Furthermore, reliable probabilistic signals can also be used for sample selection and learning regulation in large model post-training, balancing reliability and training efficiency, and driving agents to actively verify, achieving the transition from probabilistic trustworthiness to verifiable decision-making. This Tutorial follows the logic of \"recognizing the unknown—probabilistic modeling—evidence fusion—reliable decision-making,\" introducing uncertainty learning in open environments, foundations of probabilistic machine learning, reliable fusion and test-time adaptation of multimodal large models, as well as confidence calibration and risk perception in large model reasoning and agent decision-making, and further introducing reliable and efficient reasoning learning and verifiable decision-making for large models, helping participants build a knowledge framework from \"knowing that one does not know\" to \"making safe decisions in uncertain environments.\"","is_demo":0,"cover_url":"/api/forums/4030/cover?size=thumb","cover_alt":"Probabilistic Trustworthy Learning in the Era of Large Models: Uncertainty, Calibration, and Reliable Reasoning","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4030","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4030","forum_type":"专题论坛","format":"Tutorial","theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-107A三星堆A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4030","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"大模型时代的概率可信学习：不确定性、校准与可靠推理","original_topic":"未分类","original_description":"随着大模型、多模态大模型与智能体系统的快速发展，AI正从“回答问题”走向“表达推理、调用工具并参与决策”。但一个突出的问题是：大模型往往并不知道自己何时不知道——它们可能在未知问题上自信作答，在复杂推理中生成看似合理却错误的思维链，在多模态场景中受证据冲突与视觉幻觉干扰，也缺乏必要的求证、拒答与风险感知机制。可信大模型的关键已不只是提升能力规模，而是让模型识别未知、表达不确定性、校准置信度，并在不可靠时采取更安全的推理与决策策略。概率学习长期关注不确定性建模、置信度校准与可靠决策，为此提供了系统的理论与方法工具。进一步地，可靠概率信号还可用于大模型后训练中的样本选择与学习调控，兼顾可靠性与训练效率，并驱动智能体主动验证，实现从概率可信到可验证决策。本次Tutorial按照“识别未知—概率建模—证据融合—可靠决策”的逻辑，介绍开放环境中的不确定性学习、概率机器学习基础、多模态大模型的可靠融合与测试时适应，以及大模型推理与智能体决策中的置信校准与风险感知，并进一步介绍面向大模型的可靠高效推理学习与可验证决策，帮助参会者建立从“知道自己不知道”到“在不确定环境中安全决策”的知识框架。","people":[{"forum_id":4030,"role":"guest","name":"胡清华","affiliation":"天津大学","sort_order":0},{"forum_id":4030,"role":"guest","name":"张长青","affiliation":"天津大学","sort_order":1},{"forum_id":4030,"role":"guest","name":"胡文波","affiliation":"合肥工业大学","sort_order":2},{"forum_id":4030,"role":"guest","name":"韩宗博","affiliation":"北京邮电大学","sort_order":3},{"forum_id":4030,"role":"guest","name":"崔鹏","affiliation":"清华大学","sort_order":4},{"forum_id":4030,"role":"guest","name":"祝继华","affiliation":"西安交通大学","sort_order":5}],"host":"","image_url":"/api/forums/4030/cover"},{"id":4032,"title":"Time Series Foundation Models and Dynamic World Modeling","topic":"Unclassified","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-207A天府厅A","description":"With the widespread application of time series data in dynamic scenarios such as industrial systems, energy networks, urban operations, and financial markets, how to build general time series foundation models to achieve perception, prediction, and decision-making for dynamic systems has become a research hotspot. Centered on the goal of \"dynamic world modeling,\" time series foundation models need to further break through deployment bottlenecks such as robustness, interpretability, and efficiency on top of the scale foundation of automation and generality, while also exploring time series generation and reasoning capabilities to form a complete closed loop from perception to decision-making.\n\nThis tutorial aims at \"dynamic world modeling\" and uses the AGREE principles (Automation, Generalization, Robustness, Explainability, Efficiency) as its framework to systematically introduce frontier advances in time series foundation models. The content covers four main parts: (1) An overview of time series foundation models and dynamic world modeling, introducing the definition and technological evolution of time series foundation models, and reviewing the technical paths and latest advances in automation and generality; (2) Around robustness, interpretability, and efficiency, introducing multimodal time series foundation models, domain-knowledge-driven interpretable time series modeling, and model compression techniques for edge deployment; (3) Generative-mode-driven cross-domain time series modeling, introducing how generative modeling unifies heterogeneous time series data representations and improves the cross-domain generalization ability of foundation models; (4) Time series reasoning for the dynamic world, introducing understanding and reasoning methods that align time series with large language models, as well as structured deep-thinking paradigms for complex system diagnosis, and discussing the time series reasoning closed loop from perception and cognition to decision-making and action.","is_demo":0,"cover_url":"/api/forums/4032/cover?size=thumb","cover_alt":"Time Series Foundation Models and Dynamic World Modeling","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4032","source_checked_at":"2026-10-30","agenda_json":"[[\"Overview of Time Series Foundation Models and Dynamic World Modeling\",\"杨彬\"],[\"Robustness, Interpretability, and Efficiency in the AGREE Principles\",\"赵锴\"],[\"Generative Pattern-Driven Cross-Domain Time Series Modeling Methods\",\"周号益\"],[\"Time Series Reasoning for a Dynamic World\",\"裴丹\"]]","code":"CNCC2026-4032","forum_type":"专题论坛","format":"Tutorial","theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-207A天府厅A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":4,"forum_id":"CNCC2026-4032","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"时间序列基础模型和动态世界建模","original_topic":"未分类","original_description":"随着时间序列数据在工业系统、能源网络、城市运行、金融市场等动态场景中的广泛应用，如何构建通用的时序基础模型，实现对动态系统的感知、预测与决策，成为研究热点。围绕“动态世界建模”这一目标，时序基础模型需要在自动化与通用性的规模化基础之上，进一步突破鲁棒性、可解释性与高效性等部署瓶颈，同时探索时序生成与推理能力，构成从感知到决策的完整闭环。\n\n本教程旨在以“动态世界建模”为目标，以 AGREE 原则（自动化Automation、通用性Generalization、鲁棒性Robustness、可解释性Explainability、高效性Efficiency）为框架，系统介绍时序基础模型的前沿进展，内容涵盖四个主要部分：（1）时序基础模型与动态世界建模概述，介绍时序基础模型的定义与技术演进，梳理自动化与通用性的技术路径与最新进展；（2）围绕鲁棒性、可解释性与高效性，介绍时序多模态基础模型、领域知识驱动的可解释时序建模，以及面向边端部署的模型压缩技术；（3）生成模式驱动的跨域时序建模，介绍生成式建模如何统一异质时序数据表征并提升基础模型的跨域泛化能力；（4）面向动态世界的时序推理，介绍将时序与大语言模型对齐的理解与推理方法，以及面向复杂系统诊断的结构化深度思考范式，探讨从感知、认知到决策、行动的时序推理闭环。","people":[{"forum_id":4032,"role":"guest","name":"杨彬","affiliation":"华东师范大学","sort_order":0},{"forum_id":4032,"role":"guest","name":"赵锴","affiliation":"华东师范大学","sort_order":1},{"forum_id":4032,"role":"guest","name":"周号益","affiliation":"北京航空航天大学","sort_order":2},{"forum_id":4032,"role":"guest","name":"裴丹","affiliation":"清华大学","sort_order":3}],"host":"","image_url":"/api/forums/4032/cover"},{"id":4036,"title":"Agent Memory: Theory, Systems, and Applications","topic":"AI Agents","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-101海螺沟","description":"As large models evolve toward long-running agents, memory capability has become a key foundation for long-term learning, personalized services, and complex task execution. Building efficient, reliable, and scalable memory systems for large-model agents is becoming an important research direction in artificial intelligence.\nThis workshop focuses on the theoretical foundations, system architectures, and application practices of large-model memory systems, systematically introducing the development trajectory of memory technology and the latest research progress. First, it analyzes why artificial intelligence needs memory and introduces the overall architecture of modern large-model memory systems, including the constituent modules of memory, the hierarchical structure, and the complete workflow. It then focuses on different memory organization forms such as parametric memory and non-parametric memory, analyzing the characteristics, applicable scenarios, and key technologies of various memory mechanisms. On this basis, it further introduces core methods such as memory updating, memory compression, memory forgetting, and memory evolution, and explores how large models can achieve long-term continual learning and dynamic","is_demo":0,"cover_url":"/api/forums/4036/cover?size=thumb","cover_alt":"Agent Memory: Theory, Systems, and Applications","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4036","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4036","forum_type":"专题论坛","format":"Tutorial","theme_code":"Agent智能体","theme_name":"Agent智能体","venue":"天府国际会议中心","room":"天府国际会议中心-101海螺沟","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4036","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智能体记忆：理论、系统与应用","original_topic":"Agent智能体","original_description":"随着大模型向长期运行智能体演进，记忆能力已成为实现长期学习、个性化服务和复杂任务执行的关键基础。构建高效、可靠、可扩展的大模型智能体记忆系统，正成为人工智能领域的重要研究方向。\n本讲习班聚焦大模型记忆系统的理论基础、系统架构与应用实践，系统介绍记忆技术的发展脉络及最新研究进展。首先分析人工智能为什么需要记忆，并介绍现代大模型记忆系统的整体架构，包括记忆的组成模块、层次结构以及完整工作流程。随后重点讲解参数化记忆、非参数化记忆等不同记忆组织形式，分析各类记忆机制的特点、适用场景及关键技术。在此基础上，进一步介绍记忆更新、记忆压缩、记忆遗忘、记忆演化等核心方法，探讨大模型如何实现长期持续学习与动态","people":[{"forum_id":4036,"role":"guest","name":"吴乐","affiliation":"合肥工业大学","sort_order":0},{"forum_id":4036,"role":"guest","name":"陈旭","affiliation":"中国人民大学","sort_order":1},{"forum_id":4036,"role":"guest","name":"张宁豫","affiliation":"浙江大学","sort_order":2},{"forum_id":4036,"role":"guest","name":"邓淑敏","affiliation":"浙江大学","sort_order":3},{"forum_id":4036,"role":"guest","name":"薛逸达","affiliation":"浙江大学","sort_order":4},{"forum_id":4036,"role":"guest","name":"侯旻","affiliation":"合肥工业大学","sort_order":5}],"host":"","image_url":"/api/forums/4036/cover"},{"id":4037,"title":"Research Progress and Applications of Graph Foundation Models","topic":"AI Foundation Models","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-408鹿溪河","description":"As an important component of the AI data ecosystem, graphs have seen their machine learning techniques evolve from shallow methods to deep models. Meanwhile, foundation models have achieved great success in fields such as natural language processing, and their powerful emergent and generalization capabilities have aroused strong interest among graph machine learning researchers, prompting discussion on constructing the next-generation graph learning paradigm: pretraining on large-scale graph data and then adapting to various downstream graph tasks. In this workshop, we will systematically introduce the research progress and applications of \"graph foundation models.\" First, we will explain the conceptual connotations, key characteristics, and core supporting technologies of graph foundation models; then we will categorize and review research work in this direction, providing a comprehensive analysis and in-depth interpretation of the current state of development; finally, we will look ahead to future research directions and potential application prospects in this rapidly developing field.","is_demo":0,"cover_url":"/api/forums/4037/cover?size=thumb","cover_alt":"Research Progress and Applications of Graph Foundation Models","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4037","source_checked_at":"2026-10-30","agenda_json":"[[\"An Overview of Graph Foundation Models\",\"石川\"],[\"Graph Foundation Models: From Static Representation Learning to Dynamic World Modeling\",\"魏哲巍\"],[\"Graph Foundation Models: From Foundation Learning to Intelligent Applications\",\"杨成\"],[\"Graph Foundation Models: From Multi-Domain Generalization to Trustworthy Reasoning\",\"孙庆赟\"]]","code":"CNCC2026-4037","forum_type":"专题论坛","format":"Tutorial","theme_code":"AI基础模型","theme_name":"AI基础模型","venue":"天府国际会议中心","room":"天府国际会议中心-408鹿溪河","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":4,"forum_id":"CNCC2026-4037","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"图基础模型研究进展与应用","original_topic":"AI基础模型","original_description":"图作为人工智能数据生态中的重要组成部分，其机器学习技术经历了从浅层方法到深度模型的演进。与此同时，基础模型在自然语言处理等领域取得了巨大成功，其强大的涌现能力与泛化能力引发了图机器学习研究者的浓厚兴趣，促使人们开始探讨下一代图学习范式的构建：即在大规模图数据上进行预训练，进而适配于多种下游图任务。在本次讲习班中，我们将系统介绍“图基础模型”的研究进展与应用。首先阐述图基础模型的概念内涵、关键特性与核心支撑技术，随后分类梳理该方向的研究工作，对发展现状进行全面分析与深入解读，最后展望这一快速发展领域的未来研究方向与潜在应用前景。","people":[{"forum_id":4037,"role":"guest","name":"石川","affiliation":"北京邮电大学","sort_order":0},{"forum_id":4037,"role":"guest","name":"魏哲巍","affiliation":"中国人民大学","sort_order":1},{"forum_id":4037,"role":"guest","name":"杨成","affiliation":"北京邮电大学","sort_order":2},{"forum_id":4037,"role":"guest","name":"孙庆赟","affiliation":"北京航空航天大学","sort_order":3}],"host":"","image_url":"/api/forums/4037/cover"},{"id":4041,"title":"Neuro-Symbolic Learning: Theory, Methods, and Applications","topic":"Machine Learning Theory and Mechanisms","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-115A九寨厅A","description":"Neuro-symbolic learning aims to integrate data-driven machine learning with knowledge-driven symbolic reasoning. It is regarded as an important technical path for breaking through the trustworthiness and reliability bottlenecks of existing large models and realizing the next generation of artificial intelligence, and has attracted great attention from government, academia, and industry. This tutorial will systematically introduce the basic ideas, typical theoretical frameworks, representative methods, and development trajectory of neuro-symbolic learning, and will focus on discussing the new opportunities and new challenges faced by neuro-symbolic learning in the era of large models. Centering on frontier directions such as large language model reasoning enhancement, multimodal understanding and generation, AI agents, and embodied intelligence, it will deeply analyze how classical symbolic artificial intelligence methods can be deeply combined with large models, thereby improving the knowledge constraint capability, compositional generalization capability, long-range reasoning and planning capability, interpretability, and trustworthiness and reliability of artificial intelligence systems in complex open environments.","is_demo":0,"cover_url":"/api/forums/4041/cover?size=thumb","cover_alt":"Neuro-Symbolic Learning: Theory, Methods, and Applications","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4041","source_checked_at":"2026-10-30","agenda_json":"[[\"Development and Future of Neuro-Symbolic Learning\",\"戴望州\"],[\"Symbolic Logical Reasoning of Large Models\",\"李昊轩\"],[\"Neuro-Symbolic Learning-Based Multimodal Understanding and Generation\",\"丁尧相\"],[\"Neuro-Symbolic Agents\",\"郭兰哲\"]]","code":"CNCC2026-4041","forum_type":"专题论坛","format":"Tutorial","theme_code":"机器学习理论与机理","theme_name":"机器学习理论与机理","venue":"天府国际会议中心","room":"天府国际会议中心-115A九寨厅A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":4,"forum_id":"CNCC2026-4041","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"神经符号学习：理论、方法与应用","original_topic":"机器学习理论与机理","original_description":"神经符号学习旨在融合数据驱动的机器学习与知识驱动的符号推理，被认为是突破现有大模型可信可靠性瓶颈、实现下一代人工智能的重要技术途径，得到了政府、学术界、产业界的高度关注。本次Tutorial将系统介绍神经符号学习的基本思想、典型理论框架、代表性方法与发展脉络，并重点讨论在大模型时代，神经符号学习所面临的新机遇与新挑战，围绕大语言模型推理增强、多模态理解与生成、智能体、具身智能等前沿方向，深入分析经典符号主义人工智能方法如何与大模型深度结合，从而提升人工智能系统在复杂开放环境中的知识约束能力、组合泛化能力、长程推理与规划能力、可解释性和可信可靠性。","people":[{"forum_id":4041,"role":"guest","name":"戴望州","affiliation":"南京大学","sort_order":0},{"forum_id":4041,"role":"guest","name":"李昊轩","affiliation":"北京大学","sort_order":1},{"forum_id":4041,"role":"guest","name":"丁尧相","affiliation":"浙江大学","sort_order":2},{"forum_id":4041,"role":"guest","name":"郭兰哲","affiliation":"南京大学","sort_order":3}],"host":"","image_url":"/api/forums/4041/cover"},{"id":4042,"title":"Large-Scale Reinforcement Learning: From Theoretical Foundations to Large Models and Embodied Intelligence Applications","topic":"Reinforcement Learning and Reasoning","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-107B三星堆B","description":"Current AI research is in a critical period of large-model-driven emergence of digital intelligence and the rise of physical intelligence, with reinforcement learning (RL) playing a core role: large model post-training relies on RL to break through reasoning boundaries, and embodied intelligence relies on RL to achieve continuous capability evolution. This Tutorial sorts out the development trajectory of RL from classical algorithms to frontier applications. The content covers three parts: (1) theoretical foundations, including classical content such as MDP, GPI, policy and value function fitting, and the PPO algorithm, as well as frontier issues such as plasticity and scalability; (2) large model reinforcement learning, focusing on core technologies such as RLHF, post-training RL optimization, and OPD; (3) embodied intelligence reinforcement learning, discussing key issues such as RL training of embodied foundation models, online RL, and sim-to-real RL. This topic combines theoretical depth and application breadth, helping researchers grasp the underlying mechanisms of RL while providing frontier perspectives for industrial engineering practice.","is_demo":0,"cover_url":"/api/forums/4042/cover?size=thumb","cover_alt":"Large-Scale Reinforcement Learning: From Theoretical Foundations to Large Models and Embodied Intelligence Applications","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4042","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4042","forum_type":"专题论坛","format":"Tutorial","theme_code":"强化学习与推理","theme_name":"强化学习与推理","venue":"天府国际会议中心","room":"天府国际会议中心-107B三星堆B","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4042","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"大规模强化学习：从理论基础到大模型与具身智能应用","original_topic":"强化学习与推理","original_description":"当前AI研究处于大模型驱动数字智能涌现与物理智能兴起的关键时期，强化学习（RL）发挥着核心作用：大模型后训练依赖RL突破推理边界，具身智能依赖RL实现能力持续演进。本Tutorial梳理RL从经典算法到前沿应用的发展脉络。内容涵盖三部分：（1）理论基础，包括MDP、GPI、策略与值函数拟合、PPO算法等经典内容，以及可塑性、可拓展性等前沿问题；（2）大模型强化学习，聚焦 RLHF、后训练RL优化、OPD等核心技术；（3）具身智能强化学习，探讨具身基模能力RL训练、在线RL、仿真到现实RL等关键问题。本选题兼具理论深度与应用广度，既能帮助研究者把握RL底层机理，又能为产业工程时间提供前沿视角。","people":[{"forum_id":4042,"role":"guest","name":"郝建业","affiliation":"深圳忆纪元科技有限公司","sort_order":0},{"forum_id":4042,"role":"guest","name":"汤宏垚","affiliation":"天津大学","sort_order":1},{"forum_id":4042,"role":"guest","name":"毛航宇","affiliation":"中国科学院微电子研究所","sort_order":2},{"forum_id":4042,"role":"guest","name":"白辰甲","affiliation":"中国电信人工智能研究院","sort_order":3}],"host":"","image_url":"/api/forums/4042/cover"},{"id":4049,"title":"Engineering Long-Lived Intelligent Systems: Memory Architecture, Model Evolution, and Deployment","topic":"AI Agents","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-118B峨眉厅B","description":"As large models and AI agents enter cross-session, cross-task, and continuously running scenarios, the key question is no longer merely \"whether information can be recalled,\" but how to organize knowledge, experience, and state into system capabilities that are updatable, schedulable, verifiable, and governable. This tutorial takes the engineering construction of long-lived intelligent systems as its main line. The foundational theory part is used to define the capability boundaries of parametric memory, activation memory, and explicit memory; the main content focuses on system mechanisms such as memory units and hierarchical representations, write gating, retrieval scheduling, updating and forgetting, version management, and permission auditing, as well as evaluation and launch acceptance oriented toward effectiveness, efficiency, cost, and reliability. The course will combine cases such as enterprise knowledge management, long-lived agents, financial services, personal assistants, and autonomous data analysis to discuss the key trade-offs of memory systems from prototype to production environment, and further extend to the evolution of model-side memory mechanisms and memory-native architectures.","is_demo":0,"cover_url":"/api/forums/4049/cover?size=thumb","cover_alt":"Engineering Long-Lived Intelligent Systems: Memory Architecture, Model Evolution, and Deployment","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4049","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4049","forum_type":"专题论坛","format":"Tutorial","theme_code":"Agent智能体","theme_name":"Agent智能体","venue":"天府国际会议中心","room":"天府国际会议中心-118B峨眉厅B","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4049","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"长期智能系统工程：记忆架构、模型演进与部署","original_topic":"Agent智能体","original_description":"当大模型与智能体进入跨会话、跨任务和持续运行场景，关键问题已不只是“能否召回信息”，而是如何把知识、经验与状态组织成可更新、可调度、可验证、可治理的系统能力。本 Tutorial 以长期智能系统的工程化构建为主线，基础理论部分用于界定参数记忆、激活记忆与明文记忆的能力边界；主体内容聚焦记忆单元与分层表示、写入门控、检索调度、更新与遗忘、版本管理、权限审计等系统机制，以及面向效果、效率、成本和可靠性的评测与上线验收。课程将结合企业知识管理、长期智能体、金融服务、个人助手与自治数据分析等案例，讨论记忆系统从原型到生产环境的关键取舍，并进一步延伸至模型侧记忆机制和记忆原生架构的演进。","people":[{"forum_id":4049,"role":"guest","name":"王昊奋","affiliation":"同济大学","sort_order":0},{"forum_id":4049,"role":"guest","name":"李志宇","affiliation":"记忆张量（科技）有限公司","sort_order":1},{"forum_id":4049,"role":"guest","name":"周煊赫","affiliation":"上海交通大学电子信息与电气工程学院","sort_order":2}],"host":"","image_url":"/api/forums/4049/cover"},{"id":4054,"title":"AI4Science: Mechanisms, Methods, and Applications","topic":"AI4Science","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-406A兴隆湖A","description":"Currently, AI4Science is moving from single-point technological breakthroughs toward a new stage of systematic development. It has been deeply integrated into the complete scientific research process from literature reading and experimental design to scientific discovery, and has shown transformative potential in fields such as life sciences, new materials, and new energy, becoming a strategic high ground in global technological competition. Following this trend, this workshop focuses on the mechanisms, methods, and applications of AI4Science. It first reviews the development trajectory of AI4Science, deeply analyzes core concepts such as the integration of data and mechanisms and symmetry constraints, and systematically explains underlying algorithms such as graph neural networks, physics-informed neural networks, and generative models. On this basis, it introduces frontier technology platforms such as scientific large models, AI Agent autonomous reasoning, and automated closed-loop experiments combining dry and wet labs. Finally, it selects two typical scenarios, physics and life health, to deeply explore the application practices of AI in meteorological science and food science.","is_demo":0,"cover_url":"/api/forums/4054/cover?size=thumb","cover_alt":"AI4Science: Mechanisms, Methods, and Applications","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4054","source_checked_at":"2026-10-30","agenda_json":"[[\"AI4Science: History, Core Concepts, and Technical Frameworks\",\"田永鸿\"],[\"Scientific Large Models, AI Agents, and Automated Experiments\",\"江俊\"],[\"Practical Applications of AI4Science in Atmospheric Science\",\"白琮\"],[\"AI-Driven Discovery in Food Science\",\"闵巍庆\"]]","code":"CNCC2026-4054","forum_type":"专题论坛","format":"Tutorial","theme_code":"AI4Science","theme_name":"AI4Science","venue":"天府国际会议中心","room":"天府国际会议中心-406A兴隆湖A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":4,"forum_id":"CNCC2026-4054","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI4Science：机理、方法与应用","original_topic":"AI4Science","original_description":"当前，AI4Science正从单点技术突破迈向体系化发展的新阶段，已深度融入从文献阅读、实验设计到科学发现的完整科研流程，并在生命科学、新材料、新能源等领域中展现出变革性潜力，成为全球科技博弈的战略高地。顺应这一趋势，本讲习班聚焦AI4Science的机理、方法与应用，首先梳理AI4Science的发展脉络，深度剖析数据与机理融合、对称性约束等核心理念，系统讲解图神经网络、物理信息神经网络及生成式模型等底层算法。在此基础上，介绍科学大模型、AI Agent自主推理以及干湿结合的自动化闭环实验等前沿技术平台。最后，选取物理与生命健康两个典型场景，深入探讨AI在气象科学与食品科学中的应用实践。","people":[{"forum_id":4054,"role":"guest","name":"田永鸿","affiliation":"北京大学","sort_order":0},{"forum_id":4054,"role":"guest","name":"江俊","affiliation":"中国科学技术大学","sort_order":1},{"forum_id":4054,"role":"guest","name":"白琮","affiliation":"浙江工业大学","sort_order":2},{"forum_id":4054,"role":"guest","name":"闵巍庆","affiliation":"中国科学院计算技术研究所","sort_order":3}],"host":"","image_url":"/api/forums/4054/cover"},{"id":4060,"title":"Physical Intelligence: Foundations and Frontiers","topic":"Robotics and Embodied Intelligence","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-106锦城","description":"Physical intelligence is the next-generation core frontier field for artificial intelligence to break through the limitations of pure data-driven approaches, integrating physical constraints with intelligent computing, with research scope including key directions such as embodied intelligence, spatial intelligence, world models, and intelligent software-hardware co-design. This tutorial sets up two core content modules: the first part is a panoramic overview of the physical intelligence field, sorting out the disciplinary context, interdisciplinary system, and development trends; the second part breaks down the above four frontier technologies by module, systematically explaining the basic theories, mainstream technical routes, frontier research progress, and deployment pain points of each branch. The course is aimed at participating researchers and technical practitioners, providing systematic knowledge explanation to help participants build a complete physical intelligence knowledge system and accurately grasp the research hotspots and future development directions of the field.","is_demo":0,"cover_url":"/api/forums/4060/cover?size=thumb","cover_alt":"Physical Intelligence: Foundations and Frontiers","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4060","source_checked_at":"2026-10-30","agenda_json":"[[\"Overview of the Physical Intelligence Field and Key Technologies\",\"张伟楠\"],[\"World Models: Progress and Prospects\",\"沈春华\"],[\"3D Representation and Reconstruction Methods for Spatial Intelligence\",\"于沐霖\"],[\"Key Technologies and Applications of Heterogeneous Embodied Intelligence Software-Hardware Collaboration\",\"赵斌\"],[\"Continuous Improvement and Safety Constraints of Embodied Intelligence\",\"顾尚定\"]]","code":"CNCC2026-4060","forum_type":"专题论坛","format":"Tutorial","theme_code":"机器人与具身智能","theme_name":"机器人与具身智能","venue":"天府国际会议中心","room":"天府国际会议中心-106锦城","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":5,"forum_id":"CNCC2026-4060","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"物理智能：基础与前沿","original_topic":"机器人与具身智能","original_description":"物理智能是人工智能突破纯数据驱动局限的下一代核心前沿领域，融合物理约束与智能计算，研究范畴包含具身智能、空间智能、世界模型、智能软硬件协同等关键方向。本教程设置两大核心内容模块：第一部分为物理智能领域全景概览，梳理学科脉络、交叉体系与发展趋势；第二部分分模块拆解上述四大前沿技术，系统讲解各分支基础理论、主流技术路线、前沿研究进展与落地痛点。课程面向参会科研人员与技术从业者，提供系统化知识讲解，帮助参会者搭建完整物理智能知识体系，精准把握领域研究热点与未来发展方向。","people":[{"forum_id":4060,"role":"guest","name":"张伟楠","affiliation":"上海交通大学","sort_order":0},{"forum_id":4060,"role":"guest","name":"沈春华","affiliation":"浙江大学","sort_order":1},{"forum_id":4060,"role":"guest","name":"于沐霖","affiliation":"浦江实验室","sort_order":2},{"forum_id":4060,"role":"guest","name":"赵斌","affiliation":"西北工业大学","sort_order":3},{"forum_id":4060,"role":"guest","name":"顾尚定","affiliation":"上海交通大学","sort_order":4}],"host":"","image_url":"/api/forums/4060/cover"},{"id":4061,"title":"Continual Learning for Open Environments: Theory, Methods, and New Challenges in the Era of Large Models","topic":"Machine Learning Theory and Mechanisms","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-204A蜀都厅A","description":"Continual learning is an important foundation for realizing intelligent systems in open environments. It aims to enable models to continuously acquire new knowledge in constantly changing data, tasks, and scenarios while maintaining existing capabilities. As artificial intelligence systems move toward real dynamic environments, issues such as catastrophic forgetting, knowledge transfer, update cost, and long-term adaptability are becoming increasingly critical. In recent years, the development of pretrained models, vision-language models, large language models, and AI Agents has further expanded the research boundaries of continual learning. This tutorial will systematically review the basic definitions, task settings, evaluation metrics, and classical methods of continual learning, focus on introducing new paradigms in the era of pretrained models, and, in combination with issues such as knowledge updating, long-term memory, and personalized adaptation in large models, multimodal models, and AI Agents, discuss the application value, practical challenges, and future trends of continual learning.","is_demo":0,"cover_url":"/api/forums/4061/cover?size=thumb","cover_alt":"Continual Learning for Open Environments: Theory, Methods, and New Challenges in the Era of Large Models","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4061","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4061","forum_type":"专题论坛","format":"Tutorial","theme_code":"机器学习理论与机理","theme_name":"机器学习理论与机理","venue":"天府国际会议中心","room":"天府国际会议中心-204A蜀都厅A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4061","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"面向开放环境的持续学习：理论、方法与大模型时代的新挑战","original_topic":"机器学习理论与机理","original_description":"持续学习是实现开放环境智能系统的重要基础，旨在使模型在不断变化的数据、任务和场景中持续获取新知识，同时保持已有能力。随着人工智能系统走向真实动态环境，灾难性遗忘、知识迁移、更新成本和长期适应能力等问题日益关键。近年来，预训练模型、视觉语言模型、大语言模型和AI Agent的发展进一步拓展了持续学习的研究边界。本Tutorial将系统梳理持续学习的基本定义、任务设定、评价指标和经典方法，重点介绍预训练模型时代的新范式，并结合大模型、多模态模型和AI Agent中的知识更新、长期记忆、个性化适应等问题，探讨持续学习的应用价值、实践挑战与未来趋势。","people":[{"forum_id":4061,"role":"guest","name":"周大蔚","affiliation":"南京大学","sort_order":0},{"forum_id":4061,"role":"guest","name":"刘夏雷","affiliation":"南开大学","sort_order":1},{"forum_id":4061,"role":"guest","name":"郑伟诗","affiliation":"中山大学","sort_order":2},{"forum_id":4061,"role":"guest","name":"王楠楠","affiliation":"西安电子科技大学","sort_order":3},{"forum_id":4061,"role":"guest","name":"张利军","affiliation":"南京大学","sort_order":4}],"host":"","image_url":"/api/forums/4061/cover"},{"id":4065,"title":"Multi-Agent Systems: From Methods to Safety, from Evaluation to Practice","topic":"AI Agents","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-104黄龙","description":"Agent systems have been widely applied in important scenarios such as software engineering, industrial automation, and biology, becoming an important form for large models to move toward complex task solving and real-world deployment. This Tutorial will focus on three aspects—\"multi-agent methods, safety, and evaluation\"—to systematically introduce the fusion theory and collaboration mechanisms of agent systems. Subsequently, addressing the safety risks faced by agent systems in real-world deployment, it will introduce safety evaluation methods based on Agent Trajectory, as well as the full-stack agent evaluation system developed by the team—the AEP system—demonstrating the complete evaluation process from task onboarding, sandbox execution, trajectory collection, automatic evaluation, to failure attribution, helping participants comprehensively understand the theoretical foundations, key technologies, and engineering practices of agent system evaluation. The above content has attracted widespread attention from both academia and industry and is highly relevant to the CNCC audience, helping them understand the particularity and importance of agent evaluation and master the core ideas and methods of agent evaluation.","is_demo":0,"cover_url":"/api/forums/4065/cover?size=thumb","cover_alt":"Multi-Agent Systems: From Methods to Safety, from Evaluation to Practice","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4065","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4065","forum_type":"专题论坛","format":"Tutorial","theme_code":"Agent智能体","theme_name":"Agent智能体","venue":"天府国际会议中心","room":"天府国际会议中心-104黄龙","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4065","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"多智能体系统：从方法到安全，从评测到实践","original_topic":"Agent智能体","original_description":"智能体系统已经广泛应用于软件工程、工业自动化、生物学等重要场景，成为大模型走向复杂任务求解和真实场景落地的重要形态。本 Tutorial 将聚焦 “多智能体方法、安全与评测” 三个方面展，系统介绍智能体系统的融合理论和协作机制。随后，针对智能体系统面向真实世界部署所面临的安全风险，介绍基于执行轨迹（Agent Trajectory）的安全评测方法，以及团队开发的全栈智能体评测系统——智航系统（AEP），展示从任务接入、沙箱执行、轨迹采集、自动评测到失败归因的完整评测流程，帮助参会者全面理解智能体系统评测的理论基础、关键技术与工程实践。上述内容均受到学术界和产业界的广泛关注，也与 CNCC 观众高度相关，将帮助观众理解智能体评测的特殊性和重要性，掌握智能体评测的核心思想和方法。","people":[{"forum_id":4065,"role":"guest","name":"邹娜","affiliation":"上海人工智能实验室","sort_order":0},{"forum_id":4065,"role":"guest","name":"胡舒悦","affiliation":"上海人工智能实验室","sort_order":1},{"forum_id":4065,"role":"guest","name":"陈阳","affiliation":"上海人工智能实验室","sort_order":2},{"forum_id":4065,"role":"guest","name":"张鸣轩","affiliation":"上海人工智能实验室","sort_order":3},{"forum_id":4065,"role":"guest","name":"王冠楚","affiliation":"上海人工智能实验室","sort_order":4},{"forum_id":4065,"role":"guest","name":"王浩宁","affiliation":"上海人工智能实验室","sort_order":5}],"host":"","image_url":"/api/forums/4065/cover"},{"id":4074,"title":"Quantum Artificial Intelligence","topic":"Unclassified","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-406B兴隆湖B","description":"Quantum artificial intelligence is a frontier field at the deep intersection of quantum computing and artificial intelligence, aiming to reconstruct the computing paradigm of the post-Moore era through bidirectional empowerment. On the one hand, current quantum computing faces challenges in the number and quality of qubits as it moves toward practical application, and artificial intelligence is regarded as a key means to help quantum computing achieve complex system control, quantum error correction, and so on; on the other hand, quantum computing is believed to be able to elevate classical artificial intelligence to a new level of capability. This Tutorial systematically introduces quantum artificial intelligence from four dimensions: fundamentals of quantum computing, quantum algorithms, AI-empowered quantum systems, and the capabilities and boundaries of quantum computing. At the same time, authoritative experts in the quantum field are invited as guests for joint discussion and exchange.","is_demo":0,"cover_url":"/api/forums/4074/cover?size=thumb","cover_alt":"Quantum Artificial Intelligence","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4074","source_checked_at":"2026-10-30","agenda_json":"[[\"Principles and Development of Quantum Computing\",\"劳玲玲\"],[\"Towards Practical Universal Quantum Computing\",\"袁骁\"],[\"Learning Large-Scale Quantum Systems with AI\",\"吴亚东\"],[\"Capabilities and Boundaries of Quantum Computing\",\"陈然一鎏\"]]","code":"CNCC2026-4074","forum_type":"专题论坛","format":"Tutorial","theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-406B兴隆湖B","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":4,"forum_id":"CNCC2026-4074","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"量子人工智能","original_topic":"未分类","original_description":"量子人工智能是量子计算与人工智能深度交叉的前沿领域，旨在通过双向赋能重构后摩尔时代的计算范式。一方面，当前量子计算迈向实用化面临量子比特数量与质量的挑战，人工智能被认为是帮助量子计算实现复杂系统控制、量子纠错等的关键手段；另一方面，量子计算被认为能够将经典人工智能提升至新的能力高度。本Tutorial从四个维度系统性介绍量子人工智能：量子计算基础、量子算法、人工智能赋能量子系统、量子计算能力与边界。同时，邀请量子领域权威专家担任嘉宾，共同研讨交流。","people":[{"forum_id":4074,"role":"guest","name":"金海","affiliation":"华中科技大学","sort_order":0},{"forum_id":4074,"role":"guest","name":"劳玲玲","affiliation":"国防科技大学","sort_order":1},{"forum_id":4074,"role":"guest","name":"袁骁","affiliation":"北京大学","sort_order":2},{"forum_id":4074,"role":"guest","name":"吴亚东","affiliation":"上海交通大学","sort_order":3},{"forum_id":4074,"role":"guest","name":"陈然一鎏","affiliation":"粤港澳大湾区量子科学中心","sort_order":4},{"forum_id":4074,"role":"guest","name":"孙晓明","affiliation":"中国科学院计算技术研究所","sort_order":5},{"forum_id":4074,"role":"guest","name":"李绿周","affiliation":"中山大学","sort_order":6},{"forum_id":4074,"role":"guest","name":"吴俊杰","affiliation":"国防科技大学","sort_order":7},{"forum_id":4074,"role":"guest","name":"薛诗川","affiliation":"国防科技大学","sort_order":8}],"host":"","image_url":"/api/forums/4074/cover"},{"id":4075,"title":"AI-Driven Social Science: LLM High-Throughput Computing Framework and Educational Big Data Practice","topic":"AI4Science","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-204B蜀都厅B","description":"This course targets interdisciplinary research across artificial intelligence, psychology, and computational social science. Taking \"AI for Social Sciences\" as its overall framework, it systematically introduces how large language models are driving social science research to shift from small-sample measurement, manual coding, and static analysis toward scalable natural language analysis and high-throughput computing. With moral and emotion computing and educational evaluation as core cases, the course explains the social intuitionist model, moral foundations theory, and theory-driven dictionaries such as cMFD and eduMFD, and compares the principles, advantages, and limitations of three text analysis paradigms: dictionary matching, supervised learning, and generative large models. The practical part will use RateMyProfessors teaching evaluation data to demonstrate how to build a batch-capable, reproducible, and transferable high-throughput inference pipeline through Docker, GPU, and vLLM? The reason for recommending this topic is that it not only focuses on the theoretical and applied value of moral emotion computing, but also provides an engineering solution with significant speedup, which can raise inference throughput from about 0.38 token/s to about 16000 token/s, compressing annotation tasks that originally took hundreds of days into hours, truly achieving \"industrial deployment and teaching reuse\".","is_demo":0,"cover_url":"/api/forums/4075/cover?size=thumb","cover_alt":"AI-Driven Social Science: LLM High-Throughput Computing Framework and Educational Big Data Practice","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4075","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4075","forum_type":"专题论坛","format":"Tutorial","theme_code":"AI4Science","theme_name":"AI4Science","venue":"天府国际会议中心","room":"天府国际会议中心-204B蜀都厅B","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4075","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"人工智能驱动的社会科学：大模型高通量计算框架与教育大数据实践","original_topic":"AI4Science","original_description":"本课程面向人工智能、心理学与计算社会科学的交叉研究，以“AI for Social Sciences”为总体框架，系统介绍大语言模型如何推动社会科学研究由小样本测量、人工编码和静态分析，转向可扩展的自然语言分析与高通量计算。课程以道德与情绪计算、教育评价为核心案例，讲解社会直觉主义模型、道德基础理论及cMFD、eduMFD等理论驱动词典，并比较词典匹配、监督学习与生成式大模型三类文本分析范式的原理、优势与局限。实践部分将基于RateMyProfessors教学评价数据，展示如何通过 Docker、GPU与vLLM 构建可批量、可复现、可迁移的高通量推理流程？推荐此选题的理由是：它不仅关注道德情绪计算的理论和应用价值，更提供了显著提速的工程方案，可将推理吞吐从约0.38 token/s提升至约16000token/s，把原本耗时数百日的标注任务压缩到小时级，真正实现“工业部署与教学复用”。","people":[{"forum_id":4075,"role":"guest","name":"何吉波","affiliation":"南京师范大学","sort_order":0},{"forum_id":4075,"role":"guest","name":"黄逸杰","affiliation":"南京师范大学","sort_order":1},{"forum_id":4075,"role":"guest","name":"陈思羽","affiliation":"南京师范大学","sort_order":2},{"forum_id":4075,"role":"guest","name":"吴胜涛","affiliation":"吉林大学","sort_order":3}],"host":"","image_url":"/api/forums/4075/cover"},{"id":4077,"title":"Mechanism Compliance Auditing and Intrinsic Security Hardening","topic":"AI Safety and Governance","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-105A都江堰A","description":"The lack of intrinsic security mechanisms in large language models has become a core bottleneck for their trustworthy deployment in high-risk scenarios. Current mainstream AI safety research mostly focuses on external constraints at the output layer, making it difficult to guarantee the reliability of underlying mechanisms; constrained by the challenge of neuron polysemanticity, existing mechanistic interpretability techniques are difficult to support traceable and verifiable mechanism auditing.\n\nThis tutorial focuses on the equivalent interaction mechanism explanation theory, breaking through the underlying limitations of traditional methods, and combines multiple scenarios to explain in detail its applications in security auditing, intrinsic hardening, training optimization, and other directions, analyzing hidden risks of models and outlining the direction of technological evolution. The tutorial will demonstrate that neural network decisions can be decomposed into sparse interaction mechanisms, construct a formalized auditing scheme and empirically identify the embedded risk mechanisms of open-source models, decompose the dual-mechanism distribution law of generalization ability, verify that interaction mechanisms are the first-principles representation of model characteristics, and propose a mechanism-aligned training paradigm to balance efficiency and mechanism safety.","is_demo":0,"cover_url":"/api/forums/4077/cover?size=thumb","cover_alt":"Mechanism Compliance Auditing and Intrinsic Security Hardening","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4077","source_checked_at":"2026-10-30","agenda_json":"[[\"Mechanism Auditing and Mechanism Alignment: Breaking the Efficiency Bottleneck of Large Model Self-Evolution\",\"张拳石\"],[\"Rigorous Interpretation, Defect Localization, and Targeted Repair of Deep Model Representations\",\"邓辉琦\"],[\"Guarding Against Risks: From Large Models to Agent Security\",\"刘东瑞\"],[\"Data-Model-Agent: Perception—Cognition—Action—Evolution\",\"殷赵霞\"]]","code":"CNCC2026-4077","forum_type":"专题论坛","format":"Tutorial","theme_code":"AI安全与治理","theme_name":"AI安全与治理","venue":"天府国际会议中心","room":"天府国际会议中心-105A都江堰A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":4,"forum_id":"CNCC2026-4077","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"机理合规审计与内生安全加固","original_topic":"AI安全与治理","original_description":"大语言模型内生安全机理缺失，已成为其高风险场景可信落地的核心瓶颈。当前主流 AI 安全研究多聚焦输出层外部约束，难以保障底层机理可靠性；受神经元多义性难题制约，现有机理解释技术，难以支撑可追溯、可核验的机理审计。\n\n本教程聚焦等效交互机理解释理论，突破传统方法底层局限，结合多场景详解其在安全审计、内生加固、训练优化等方向的应用，剖析模型隐性风险并梳理技术演进方向。教程将证明神经网络决策可解构为稀疏交互机理，构建形式化审计方案并实证开源模型内嵌风险机理，拆解泛化能力的双机理分布规律，验证交互机理是模型特性第一性表征，提出机理对齐训练范式以兼顾效率与机理安全。","people":[{"forum_id":4077,"role":"guest","name":"张拳石","affiliation":"上海交通大学","sort_order":0},{"forum_id":4077,"role":"guest","name":"邓辉琦","affiliation":"西安交通大学","sort_order":1},{"forum_id":4077,"role":"guest","name":"刘东瑞","affiliation":"上海人工智能实验室","sort_order":2},{"forum_id":4077,"role":"guest","name":"殷赵霞","affiliation":"华东师范大学","sort_order":3}],"host":"","image_url":"/api/forums/4077/cover"},{"id":4087,"title":"Aerospace Embodied Intelligence: Low-Altitude UAV Navigation and Spacecraft On-Orbit Operations","topic":"Robotics and Embodied Intelligence","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-114瓦屋山","description":"This Tutorial focuses on aerospace embodied intelligence, centering on low-altitude UAVs, spacecraft on-orbit operations, and data computing support, and discusses the key issues of embodied intelligence moving from ground scenarios to open aerospace environments. The course will introduce the fundamentals of embodied intelligence and the current state of ground system development, and analyze the special constraints of aerospace scenarios in terms of complex dynamics, environmental uncertainty, communication limitations, and real-time control; further, in combination with low-altitude UAV navigation, complex maneuvering, mission planning, and autonomous decision-making in open environments, it will discuss the supporting role of embodied intelligence; and, oriented toward spacecraft and on-orbit robotic arm operations, it will explain the adaptation requirements of existing models in perception, planning, control, simulation verification, and reliable deployment. This topic aligns with the development needs of the low-altitude economy, commercial aerospace, and on-orbit services, and has important academic value and application prospects.","is_demo":0,"cover_url":"/api/forums/4087/cover?size=thumb","cover_alt":"Aerospace Embodied Intelligence: Low-Altitude UAV Navigation and Spacecraft On-Orbit Operations","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4087","source_checked_at":"2026-10-30","agenda_json":"[[\"Overall Framework of Aerospace Embodied Intelligence and UAV Decision-Making and Planning System in Open Environments\",\"刘偲\"],[\"UAV Embodied Perception and Vision-Language Navigation\",\"王鹏\"],[\"Preliminary Exploration of Key Technologies and Applications of Space Cloud Intelligence\",\"孙显\"],[\"Frontier Technologies and Applications of Near-Space Physical Intelligence\",\"袁野\"]]","code":"CNCC2026-4087","forum_type":"专题论坛","format":"Tutorial","theme_code":"机器人与具身智能","theme_name":"机器人与具身智能","venue":"天府国际会议中心","room":"天府国际会议中心-114瓦屋山","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":4,"forum_id":"CNCC2026-4087","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"空天具身智能：低空无人机导航与航天在轨操作","original_topic":"机器人与具身智能","original_description":"本 Tutorial 聚焦空天具身智能，围绕低空无人机、航天器在轨操作及数据计算支撑，讨论具身智能从地面场景走向开放空天环境的关键问题。课程将介绍具身智能基础与地面系统发展现状，分析空天场景在复杂动力学、环境不确定性、通信受限和实时控制等方面的特殊约束；进一步结合低空无人机导航、复杂机动、任务规划和开放环境自主决策，探讨具身智能的支撑作用；并面向航天器及在轨机械臂操作，说明现有模型在感知、规划、控制、仿真验证和可靠部署中的适配需求。本选题契合低空经济、商业航天和在轨服务发展需求，具有重要学术价值和应用前景。","people":[{"forum_id":4087,"role":"guest","name":"刘偲","affiliation":"北京航空航天大学","sort_order":0},{"forum_id":4087,"role":"guest","name":"王鹏","affiliation":"西北工业大学","sort_order":1},{"forum_id":4087,"role":"guest","name":"孙显","affiliation":"中国科学院空天信息创新研究院","sort_order":2},{"forum_id":4087,"role":"guest","name":"袁野","affiliation":"北京理工大学","sort_order":3}],"host":"","image_url":"/api/forums/4087/cover"},{"id":4088,"title":"Architecture, Key Technologies, and AI Agent Network Evolution for 6G AI-RAN","topic":"Unclassified","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-109自流井","description":"With the rapid development of AI technology, 6G wireless networks are evolving from traditional architectures centered on connectivity and transmission into AI-native networks with capabilities for perception, reasoning, collaboration, and autonomous decision-making. This tutorial aims to systematically introduce the architecture, key technologies, and typical applications for 6G AI-RAN. Centered on three main lines—semantic intelligence, the Token mechanism, and AI agent networks—it systematically explains how future wireless networks will evolve from bit-transmission infrastructure into a new cognitive platform that supports AI agent collaboration and embodied intelligence, and includes a live demonstration of the team's self-developed Agent AI-RAN prototype system. This tutorial aims to help researchers and engineering technical personnel fully understand the architecture, key technologies, and future development directions of 6G AI-native networks, and to provide a systematic technical reference for research and industrial implementation of next-generation intelligent wireless networks.","is_demo":0,"cover_url":"/api/forums/4088/cover?size=thumb","cover_alt":"Architecture, Key Technologies, and AI Agent Network Evolution for 6G AI-RAN","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4088","source_checked_at":"2026-10-30","agenda_json":"[[\"Toward Future 6G AI-Native Networks: Basic Concepts, Key Technologies, and Outlook\",\"Tony Quek\"],[\"Agent Communication in AI-RAN: Autonomous Planning, Dynamic Collaboration, and Semantic Transmission\",\"郭帅帅\"],[\"Cognition-Empowered Agent RAN: Architecture, Memory, and Interpretable Decision-Making\",\"王玺钧\"],[\"Semantic Intelligence Empowering Agent RAN: Toward Task-Oriented AI-Native Wireless Networks\",\"冯晨远\"]]","code":"CNCC2026-4088","forum_type":"专题论坛","format":"Tutorial","theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-109自流井","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":4,"forum_id":"CNCC2026-4088","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"面向6G AI-RAN的体系架构、关键技术与智能体网络演进","original_topic":"未分类","original_description":"随着AI技术的快速发展，6G无线网络正从以连接和传输为中心的传统架构，演进为具备感知、推理、协同和自主决策能力的AI原生网络。本教程旨在系统介绍面向6G AI-RAN的体系架构、关键技术与典型应用，围绕语义智能、Token机制和智能体网络三条主线，系统阐述未来无线网络如何从比特传输基础设施演进为支撑智能体协同和具身智能的新型认知平台，并结合团队自主研发的Agent AI-RAN原型系统进行现场演示。本教程旨在帮助研究人员和工程技术人员全面理解6G AI原生网络的体系架构、关键技术和未来发展方向，为下一代智能无线网络的研究与产业落地提供系统性的技术参考。","people":[{"forum_id":4088,"role":"guest","name":"Tony Quek","affiliation":"新加坡科技设计大学","sort_order":0},{"forum_id":4088,"role":"guest","name":"郭帅帅","affiliation":"山东大学","sort_order":1},{"forum_id":4088,"role":"guest","name":"王玺钧","affiliation":"中山大学","sort_order":2},{"forum_id":4088,"role":"guest","name":"冯晨远","affiliation":"英国埃克塞特大学","sort_order":3},{"forum_id":4088,"role":"guest","name":"毛俊","affiliation":"北京触点互动信息技术有限公司","sort_order":4},{"forum_id":4088,"role":"guest","name":"杨浩","affiliation":"浙江大学","sort_order":5},{"forum_id":4088,"role":"guest","name":"王超帆","affiliation":"温州大学","sort_order":6},{"forum_id":4088,"role":"guest","name":"夏芳","affiliation":"北京触点互动信息技术有限公司","sort_order":7}],"host":"","image_url":"/api/forums/4088/cover"},{"id":4090,"title":"Key Technologies of Multimodal Large Models for the Language Education Industry","topic":"AI for Industry","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-407广都","description":"This tutorial systematically reviews the evolution from rule-based systems to multimodal large models (MLLM) empowering language education, and deeply explores key application technologies such as multimodal alignment, intelligent grading, immersive interaction, and dynamic knowledge tracing. It aims to investigate how multimodal large models drive the educational paradigm shift toward \"learning as a service\" and enable large-scale deep personalized learning, while also considering how to enhance teacher effectiveness through human-machine collaboration amid the trend of automation. Standing at the forefront of AI reshaping education, this tutorial not only reveals the enormous potential of multimodal technologies to redefine the teaching landscape, but also provides a panoramic implementation guide from underlying resources to evaluation metrics for achieving intelligent, personalized educational transformation.","is_demo":0,"cover_url":"/api/forums/4090/cover?size=thumb","cover_alt":"Key Technologies of Multimodal Large Models for the Language Education Industry","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4090","source_checked_at":"2026-10-30","agenda_json":"[[\"Foundations of Multimodal Large Models and Core Technologies for Affective Interaction\",\"王龑\"],[\"Multimodal Teaching Support Empowering Language Educators\",\"兰韵诗\"],[\"Multimodal Tutoring Technologies Empowering Language Learners\",\"王春阳\"],[\"Application Practices of Multimodal Large Models in the Language Education Industry\",\"钱卫宁\"]]","code":"CNCC2026-4090","forum_type":"专题论坛","format":"Tutorial","theme_code":"AI+行业","theme_name":"AI+行业","venue":"天府国际会议中心","room":"天府国际会议中心-407广都","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":4,"forum_id":"CNCC2026-4090","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"面向语言教育行业的多模态大模型关键技术","original_topic":"AI+行业","original_description":"本教程系统梳理从规则系统到多模态大模型（MLLM）赋能语言教育的演进历程，深度探讨了多模态对齐、智能批改、沉浸式交互及动态知识追踪等关键应用技术，旨在探究多模态大模型如何驱动教育范式转向“学习即服务”并实现规模化的深度个性化学习，同时思考如何在自动化趋势下通过人机协同增强教师效能。本教程站在人工智能重塑教育的最前沿，不仅揭示多模态技术重新定义教学格局的巨大潜力，更为实现智能化、个性化教育转型提供了从底层资源到评估指标的全景式落地指南。","people":[{"forum_id":4090,"role":"guest","name":"王龑","affiliation":"华东师范大学","sort_order":0},{"forum_id":4090,"role":"guest","name":"兰韵诗","affiliation":"华东师范大学","sort_order":1},{"forum_id":4090,"role":"guest","name":"王春阳","affiliation":"华东师范大学","sort_order":2},{"forum_id":4090,"role":"guest","name":"钱卫宁","affiliation":"华东师范大学","sort_order":3}],"host":"","image_url":"/api/forums/4090/cover"},{"id":4091,"title":"Federated Large Models and Continual Learning","topic":"Machine Learning Theory and Mechanisms","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-110卧龙","description":"In recent years, federated learning, large language models, and continual learning technologies have flourished, jointly driving the AI training paradigm to evolve from traditional centralized architectures toward distributed systems that are privacy-secure, dynamically adaptive, and multi-party collaborative. This Tutorial series will systematically analyze the key technologies of federated large models and continual learning, and deeply explore their cutting-edge applications and challenges. The specific content includes: the basic framework and parameter-efficient fine-tuning of federated large models, privacy-preserving collaborative training and cross-client knowledge aggregation methods; catastrophic forgetting, knowledge retention, and dynamic update mechanisms in continual learning; practical applications of federated continual learning in scenarios such as natural language processing, computer vision, healthcare, finance, and on-device intelligence, as well as key challenges and future opportunities faced in the implementation of federated large models, such as data heterogeneity, communication efficiency, security and trustworthiness, and continuous evolution.","is_demo":0,"cover_url":"/api/forums/4091/cover?size=thumb","cover_alt":"Federated Large Models and Continual Learning","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4091","source_checked_at":"2026-10-30","agenda_json":"[[\"Federated Learning and Federated Large Models\",\"杨强\"],[\"Federated Intelligent Computing: From Big Data to Large Models\",\"童咏昕\"],[\"Federated Large Models and Continual Learning\",\"叶茫\"],[\"Federated Continual Learning\",\"杨新\"]]","code":"CNCC2026-4091","forum_type":"专题论坛","format":"Tutorial","theme_code":"机器学习理论与机理","theme_name":"机器学习理论与机理","venue":"天府国际会议中心","room":"天府国际会议中心-110卧龙","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":4,"forum_id":"CNCC2026-4091","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"联邦大模型与持续学习","original_topic":"机器学习理论与机理","original_description":"近年来，联邦学习、大语言模型与持续学习技术蓬勃发展，共同推动人工智能训练范式由传统集中式架构，向隐私安全、动态自适应、多方协同的分布式体系演进。本系列Tutorial将系统性地解析联邦大模型与持续学习的关键技术，并深入探讨其前沿应用与挑战，具体涵盖内容包括：联邦大模型基础框架与参数高效微调、隐私保护协同训练与跨客户端知识聚合方法；持续学习中的灾难性遗忘、知识保持与动态更新机制；联邦持续学习在自然语言处理、计算机视觉、医疗、金融和端侧智能等场景中的实践应用，以及联邦大模型落地过程中面临的数据异构、通信效率、安全可信与持续演化等关键挑战和未来机遇。","people":[{"forum_id":4091,"role":"guest","name":"杨强","affiliation":"香港理工大学","sort_order":0},{"forum_id":4091,"role":"guest","name":"童咏昕","affiliation":"北京航空航天大学","sort_order":1},{"forum_id":4091,"role":"guest","name":"叶茫","affiliation":"武汉大学","sort_order":2},{"forum_id":4091,"role":"guest","name":"杨新","affiliation":"西南财经大学","sort_order":3}],"host":"","image_url":"/api/forums/4091/cover"},{"id":4244,"title":"Self-Supervised Learning in the Era of Foundation Models: Core Methods, Theoretical Principles, and Frontier Directions","topic":"AI Foundation Models","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-118A峨眉厅A","description":"In recent years, foundation models have been rapidly changing the development paradigm of machine learning and have achieved remarkable progress in fields such as natural language processing, computer vision, and multimodal learning. Self-supervised learning is the core technology supporting this transformation: models are pretrained on massive unlabeled data to learn general representations and foundational capabilities, and are then adapted to specific tasks through fine-tuning or instruction tuning. This \"pretraining–fine-tuning\" paradigm differs from the empirical risk minimization framework based on the i.i.d. assumption in traditional machine learning, and it also brings new theoretical problems and methodological challenges. This tutorial will center on \"self-supervised learning in the era of foundation models\" and systematically introduce representative methods, theoretical foundations, and frontier directions of self-supervised learning.","is_demo":0,"cover_url":"/api/forums/4244/cover?size=thumb","cover_alt":"Self-Supervised Learning in the Era of Foundation Models: Core Methods, Theoretical Principles, and Frontier Directions","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4244","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4244","forum_type":"专题论坛","format":"Tutorial","theme_code":"AI基础模型","theme_name":"AI基础模型","venue":"天府国际会议中心","room":"天府国际会议中心-118A峨眉厅A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4244","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"基础模型时代下的自监督学习：核心方法、理论原理与前沿方向","original_topic":"AI基础模型","original_description":"近年来，基础模型正在快速改变机器学习的发展范式，并在自然语言处理、计算机视觉、多模态学习等领域取得了显著进展。自监督学习是支撑这一变革的核心技术：模型通过在海量无标注数据上进行预训练，学习通用表示与基础能力，再通过微调或指令调优适配具体任务。这一“预训练—微调”范式不同于传统机器学习中基于同分布假设的经验风险最小化框架，也带来了新的理论问题和方法挑战。本Tutorial将围绕“基础模型时代下的自监督学习”展开，系统介绍自监督学习的代表性方法、理论基础与前沿方向。","people":[{"forum_id":4244,"role":"guest","name":"王奕森","affiliation":"北京大学","sort_order":0},{"forum_id":4244,"role":"guest","name":"章祺","affiliation":"北京大学","sort_order":1}],"host":"","image_url":"/api/forums/4244/cover"},{"id":4246,"title":"Diffusion Models: Foundations and Frontiers","topic":"Vision, Multimodality and World Models","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-115B九寨厅B","description":"Diffusion models are one of the two pillar architectures behind generative large models. They are widely used in computer vision and world models, and new progress has also been made in large language models. This tutorial will introduce in detail the basic principles of diffusion models, including their probabilistic formulation, training, generation algorithms, and scalable architectures. On this basis, it will introduce frontier topics such as large-scale training techniques, image and video generation models, and diffusion-based world models. Finally, it will introduce frontier advances such as diffusion language models.\n\nAs one of the foundational methods and technologies of generative modeling, diffusion models have enormous application potential in the broader field of computer science. In addition, pretrained large-scale diffusion models, such as the Sora series and the LLaDA series, are also expected to be directly applied to related fields. Compared with large language models, diffusion models have not yet been fully disseminated in the computer field. Therefore, this tutorial should have a broad base of CNCC attendees.","is_demo":0,"cover_url":"/api/forums/4246/cover?size=thumb","cover_alt":"Diffusion Models: Foundations and Frontiers","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4246","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4246","forum_type":"专题论坛","format":"Tutorial","theme_code":"视觉、多模态与世界模型","theme_name":"视觉、多模态与世界模型","venue":"天府国际会议中心","room":"天府国际会议中心-115B九寨厅B","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4246","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"扩散模型：基础与前沿","original_topic":"视觉、多模态与世界模型","original_description":"扩散模型是生成式大模型背后两大支柱架构之一，在计算机视觉、世界模型方面有广泛应用，在大语言模型方面也有新进展。本次讲习班将详细介绍扩散模型的基本原理，包括其概率表示、训练、生成算法以及可扩展架构。在此基础上，介绍大规模训练技术，图像和视频生成模型，以及基于扩散的世界模型等前沿话题。最后，介绍扩散语言模型等前沿进展。\n\n作为生成式建模的基础方法与技术之一，扩散模型在更为广泛的计算机领域有着巨大的应用潜力。此外，预训练的大规模扩散模型，如Sora系列、LLaDA系列，也有望直接应用于相关领域。而相较于大语言模型，扩散模型在计算机领域还未被充分传播。因此，本次讲习班应该具有广泛的CNCC听众基础。","people":[{"forum_id":4246,"role":"guest","name":"李崇轩","affiliation":"中国人民大学","sort_order":0},{"forum_id":4246,"role":"guest","name":"邓志杰","affiliation":"上海交通大学","sort_order":1},{"forum_id":4246,"role":"guest","name":"赵敏","affiliation":"南京大学","sort_order":2}],"host":"","image_url":"/api/forums/4246/cover"},{"id":4256,"title":"Reinforcement Learning: Theory, Methods, and Applications","topic":"Reinforcement Learning and Reasoning","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-207B天府厅B","description":"Reinforcement learning is one of the core research directions of machine learning and has recently attracted much attention due to its application in large models. This workshop will start from the fundamentals of reinforcement learning, cover deep reinforcement learning, imitation learning, multi-agent reinforcement learning, and other content, and introduce the application support of reinforcement learning in large models and embodied intelligence scenarios.","is_demo":0,"cover_url":"/api/forums/4256/cover?size=thumb","cover_alt":"Reinforcement Learning: Theory, Methods, and Applications","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4256","source_checked_at":"2026-10-30","agenda_json":"[[\"Introduction to Reinforcement Learning\",\"俞扬\"],[\"Reinforcement Learning and Imitation Learning for Large Models\",\"许天\"]]","code":"CNCC2026-4256","forum_type":"专题论坛","format":"Tutorial","theme_code":"强化学习与推理","theme_name":"强化学习与推理","venue":"天府国际会议中心","room":"天府国际会议中心-207B天府厅B","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":2,"forum_id":"CNCC2026-4256","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"强化学习：理论、方法与应用","original_topic":"强化学习与推理","original_description":"强化学习是机器学习的核心研究方向之一，近期因在大模型中的应用而备受关注。本次讲习班将从强化学习的基础知识入手，覆盖深度强化学习、模仿学习、多智能体强化学习等内容，并介绍强化学习在大模型、具身智能场景中的应用支撑。","people":[{"forum_id":4256,"role":"guest","name":"俞扬","affiliation":"南京大学","sort_order":0},{"forum_id":4256,"role":"guest","name":"袁雷","affiliation":"南京大学","sort_order":1},{"forum_id":4256,"role":"guest","name":"许天","affiliation":"南京大学","sort_order":2}],"host":"","image_url":"/api/forums/4256/cover"},{"id":4281,"title":"Toward General Modeling of Structured Data: Foundations, Frontiers, and Applications","topic":"Data Science and Data Spaces","kind":"专题论坛","chair":"待补充","starts_at":"2026-10-21T13:30:00+08:00","ends_at":"2026-10-21T17:30:00+08:00","place":"天府国际会议中心-103剑门关","description":"Structured data, represented by tabular data, time series, and graph data, underpins many core data mining applications, including recommender systems, predictive analytics, and user behavior analysis. Although traditional methods have achieved good results, they are often designed for a single task or a single dataset and lack the ability to generalize across diverse structured data problems.\n\nIn recent years, the rise of foundation models and in-context learning is reshaping the technical path of structured data modeling: from a paradigm relying on dataset-specific training to a more general, flexible reasoning paradigm with cross-task generalization capability. Recently emerging foundation models for structured data show that, through large-scale pretraining and ICL-based reasoning mechanisms, the core ideas of foundation models can be effectively extended to the structured data domain, thereby supporting capabilities such as zero-shot reasoning and cross-scenario knowledge transfer. This direction provides a new perspective for structured data modeling.","is_demo":0,"cover_url":"/api/forums/4281/cover?size=thumb","cover_alt":"Toward General Modeling of Structured Data: Foundations, Frontiers, and Applications","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4281","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4281","forum_type":"专题论坛","format":"Tutorial","theme_code":"数据科学与数据空间","theme_name":"数据科学与数据空间","venue":"天府国际会议中心","room":"天府国际会议中心-103剑门关","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[\"官网仅列嘉宾，未单独标注主席；保留嘉宾，不推定主席。\"]","agenda_count":0,"forum_id":"CNCC2026-4281","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"迈向结构化数据的通用建模：基础、前沿与应用","original_topic":"数据科学与数据空间","original_description":"以表格数据、时间序列和图数据为代表的结构化数据支撑着许多核心的数据挖掘应用，包括推荐系统、预测分析和用户行为分析等。传统方法虽然已经取得了不错的效果，但是其往往针对单一任务或单一数据集设计的，缺乏在多样化结构化数据问题之间进行泛化的能力。\n\n近年来基础模型与上下文学习的兴起，正在重塑结构化数据建模的技术路径：从依赖数据集特定训练的范式，转向更加通用、灵活且具备跨任务泛化能力的推理范式。近期涌现的结构化数据基础模型表明，通过大规模预训练以及基于 ICL 的推理机制，可以将基础模型的核心思想有效扩展至结构化数据领域，从而支持零样本推理及跨场景知识迁移等能力。此方向为结构化数据建模提供了全新视角。","people":[{"forum_id":4281,"role":"guest","name":"崔鹏","affiliation":"清华大学","sort_order":0},{"forum_id":4281,"role":"guest","name":"张兴璇","affiliation":"稳准智能（雄安）科技有限公司","sort_order":1},{"forum_id":4281,"role":"guest","name":"王啸","affiliation":"北京航空航天大学","sort_order":2},{"forum_id":4281,"role":"guest","name":"况琨","affiliation":"浙江大学","sort_order":3}],"host":"","image_url":"/api/forums/4281/cover"},{"id":4297,"title":"Plenary Forum (Oct 22) — Security Challenges of AI Agents","topic":"Unclassified","kind":"大会论坛","chair":"任奎（主持）","starts_at":"2026-10-22T11:10:00+08:00","ends_at":"2026-10-22T12:10:00+08:00","place":"西博城1号馆","description":"Large-model-driven AI agents are moving from \"able to converse\" to \"able to perceive, think, and act,\" accelerating deployment in finance, government, industry, and endpoint scenarios. With the release of autonomous planning and tool-calling capabilities, agents inherit model- and data-layer risks such as jailbreak attacks, prompt injection, and data poisoning, while also giving rise to new hazards such as memory contamination, privilege and tool misuse, protocol and supply chain attacks, and risk contagion among multi-agent systems, which are amplified in chains within the \"perception-decision-action\" loop; the traditional security boundary system is facing reconstruction with agents as the acting subject. This forum brings together experts from universities, research institutions, and leading enterprises to explore technical paths and governance frameworks for making agents \"usable, trustworthy, and controllable,\" providing theoretical support and engineering references for building the agent security ecosystem.","is_demo":0,"cover_url":"/api/forums/4297/cover?size=thumb","cover_alt":"Plenary Forum (Oct 22) — Security Challenges of AI Agents","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4297","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4297","forum_type":"大会论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城1号馆","room":"西博城1号馆","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[\"官网标注主持，未标注主席；主持与主席分开存储。\"]","agenda_count":0,"forum_id":"CNCC2026-4297","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"22日大会论坛——智能体的安全挑战","original_topic":"未分类","original_description":"大模型驱动的 AI 智能体正从“能对话”走向“能感知、会思考、可行动”，在金融、政务、工业与终端场景加速落地。随着自主规划与工具调用能力的释放，智能体既继承了越狱攻击、提示注入、数据投毒等模型与数据层风险，又衍生出记忆污染、权限与工具滥用、协议与供应链攻击、多智能体风险传染等新型隐患，并在“感知—决策—行动”闭环中被链式放大；传统安全边界体系，正面临以智能体为行为主体的重构。本论坛汇聚高校、科研机构与领军企业的专家，共探智能体“可用、可信、可控”的技术路径与治理框架，为智能体安全生态建设提供理论支撑与工程借鉴。","people":[{"forum_id":4297,"role":"host","name":"任奎","affiliation":"浙江大学","sort_order":0},{"forum_id":4297,"role":"guest","name":"于锐","affiliation":"公安部第一研究所","sort_order":0},{"forum_id":4297,"role":"guest","name":"徐恪","affiliation":"清华大学","sort_order":1},{"forum_id":4297,"role":"guest","name":"胡侠","affiliation":"上海人工智能实验室","sort_order":2},{"forum_id":4297,"role":"guest","name":"韦韬","affiliation":"蚂蚁密算科技有限公司","sort_order":3},{"forum_id":4297,"role":"guest","name":"肖亚军","affiliation":"华为","sort_order":4}],"host":"任奎","image_url":"/api/forums/4297/cover"},{"id":4005,"title":"Opportunities and Challenges of Deploying AI Agents in Real Industrial Scenarios","topic":"AI Agents","kind":"专题论坛","chair":"刘亚彬","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城2号馆-201","description":"As large model technology penetrates deeper into industry, multi-agent systems are becoming a core variable in reconstructing the paradigm of automotive design and development. This forum, jointly initiated by IAT and Yuanmu Intelligence, focuses on \"the opportunities and challenges of deploying AI Agents in real industrial scenarios,\" shares engineering practices based on agents, and explores how multi-agent systems can connect the closed-loop collaboration of automotive styling generation, CAE simulation verification, and knowledge accumulation.\nAutomotive R&D is one of the most complex scenarios for the integration of industrial software and AI. This topic fills the forum gap on \"the deployment of AI Agents in real production lines of high-end manufacturing,\" aligns with the CNCC2026 theme \"Data Empowers Industry, Intelligent Computing Shapes the Future,\" and has empirical support from IAT's complete product matrix, providing reusable technical paths and paradigm references for scholars in the computing field and industry.","is_demo":0,"cover_url":"/api/forums/4005/cover?size=thumb","cover_alt":"Opportunities and Challenges of Deploying AI Agents in Real Industrial Scenarios","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4005","source_checked_at":"2026-10-30","agenda_json":"[[\"From Design Assistance to Engineering Execution: Exploration and Practice of Automotive R&D Agents\",\"刘亚彬\"],[\"Research on Intelligent Modeling for Industrial CAD: From Visual Understanding to Design Generation\",\"薛向阳\"],[\"Baidu Famou: A Self-Evolving Super Agent\",\"李安南\"],[\"Industrial AI-Driven Intelligent Design and Industrial Upgrading\",\"赵昊\"],[\"Opportunities and Challenges in Deploying Industrial Agents\",\"姜昌浩\"],[\"Key Technologies of Large and Small Models for Intelligent Manufacturing\",\"仝晓萌\"]]","code":"CNCC2026-4005","forum_type":"专题论坛","format":null,"theme_code":"Agent智能体","theme_name":"Agent智能体","venue":"西博城2号馆","room":"西博城2号馆-201","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4005","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智能体在工业真实场景落地的机遇与挑战","original_topic":"Agent智能体","original_description":"随着大模型技术向工业深水区渗透，多智能体系统正成为重构汽车设计研发范式的核心变量。本论坛由阿尔特与圆木智能联合发起，聚焦\"AI Agent在工业真实场景落地的机遇与挑战\"，分享基于智能体的工程化实践，探讨多Agent如何贯通汽车造型生成、CAE仿真验证与知识沉淀的闭环协同。\n汽车研发是工业软件与AI融合的最复杂场景之一。本选题填补了\"AI Agent在高端制造业真实产线落地\"的论坛空白，既契合CNCC2026\"数聚产业·智算未来\"主题，又具备阿尔特完整产品矩阵的实证支撑，可为计算领域学者与工业界提供可复用的技术路径与范式参考。","people":[{"forum_id":4005,"role":"chair","name":"刘亚彬","affiliation":"阿尔特太乙人工智能","sort_order":0},{"forum_id":4005,"role":"guest","name":"薛向阳","affiliation":"复旦大学 计算机科学技术学院","sort_order":0},{"forum_id":4005,"role":"guest","name":"李安南","affiliation":"百度","sort_order":1},{"forum_id":4005,"role":"guest","name":"赵昊","affiliation":"清华大学","sort_order":2},{"forum_id":4005,"role":"guest","name":"姜昌浩","affiliation":"圆木智能科技(北京)有限公司","sort_order":3},{"forum_id":4005,"role":"guest","name":"仝晓萌","affiliation":"北京航空航天大学","sort_order":4}],"host":"","image_url":"/api/forums/4005/cover"},{"id":4012,"title":"Enterprise Agent Practice from \"Demo\" to \"Production\"","topic":"Unclassified","kind":"专题论坛","chair":"茹炳晟、黄佳","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城2号馆-206","description":"Cool demos are all alike; the Agent that can truly withstand production-grade tests is one in ten thousand.\nCurrently, AI Agents are standing at the crossroads from \"proof of concept\" to \"value implementation.\" This forum is designed precisely for those pioneers who are unwilling to merely do demonstrations and are eager to tackle real business challenges.\nWe will focus on the most difficult and critical gaps in enterprise-level Agent implementation: How to break through context length limits? How to build stable and reliable memory and planning capabilities? And how to balance cost and latency under massive calls? There are no exaggerated PPTs here, only frontline practical code, architecture selection pitfalls, and hardcore performance tuning insights.","is_demo":0,"cover_url":"/api/forums/4012/cover?size=thumb","cover_alt":"Enterprise Agent Practice from \"Demo\" to \"Production\"","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4012","source_checked_at":"2026-10-30","agenda_json":"[[\"Agent Implementation Practices with the ADPS Agent Design Pattern\",\"黄佳\"],[\"How Organizations Implement Agents in the \\\"AGI Era\\\"\",\"weimin wang\"],[\"Enterprise Agent Practices in the GIS Ecosystem: From \\\"Demo\\\" to \\\"Production\\\"\",\"熊钰柯\"],[\"Implementation Paradigms for \\\"Specialized and Refined\\\" Small Agents on the Production Side Across Multiple Scenarios and Industries\",\"邓光磊\"],[\"Industrial-Grade Multi-Agent Architecture: Collaborative Evolution Based on \\\"Pass-by-Reference\\\" and \\\"Event-Driven\\\" Approaches\",\"范海燕\"]]","code":"CNCC2026-4012","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城2号馆","room":"西博城2号馆-206","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4012","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"企业级Agent实践从“Demo”到“生产”","original_topic":"未分类","original_description":"炫酷的Demo千篇一律，真正能扛住生产级考验的Agent万里挑一。\n当前，AI Agent正站在从“概念验证”走向“价值落地”的十字路口。这场论坛，正是为那些不甘于只做演示、渴望攻克真实业务难题的先行者而设。\n我们将聚焦企业级Agent落地中最棘手、最关键的鸿沟：如何突破上下文长度限制？如何构建稳定可靠的记忆与规划能力？又如何在海量调用下平衡成本与延迟？这里没有天花乱坠的PPT，只有来自一线的实战代码、架构选型踩坑，以及性能调优的硬核干货。","people":[{"forum_id":4012,"role":"chair","name":"茹炳晟","affiliation":"腾讯","sort_order":0},{"forum_id":4012,"role":"guest","name":"weimin wang","affiliation":"百融智能","sort_order":0},{"forum_id":4012,"role":"co_chair","name":"黄佳","affiliation":"新加坡科技研究局","sort_order":1},{"forum_id":4012,"role":"guest","name":"熊钰柯","affiliation":"成都玄宿科技有限公司","sort_order":1},{"forum_id":4012,"role":"guest","name":"邓光磊","affiliation":"四川迈作科技有限公司","sort_order":2},{"forum_id":4012,"role":"guest","name":"范海燕","affiliation":"万联智链","sort_order":3},{"forum_id":4012,"role":"guest","name":"牟青秀","affiliation":"百融智能","sort_order":4}],"host":"","image_url":"/api/forums/4012/cover"},{"id":4013,"title":"13th Frontier Forum on Intelligent Information Systems: Data Agents Empowering Industrial Digital-Intelligent Innovation","topic":"AI Agents","kind":"专题论坛","chair":"邢春晓","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-410","description":"Large-model technology continues to drive the iterative upgrading of intelligent information systems. The Data Agent is a core technology of the new generation of data intelligence; it can address the pain points of inefficient traditional data analysis and insufficient value release, and is a frontier hotspot in the field. This forum focuses on the innovative empowerment of Data Agents, conducting thematic discussions around core technologies and industrial applications. Relying on large-model semantic understanding and intelligent orchestration, Data Agents realize full-process autonomy in data processing and decision-making, make up for the shortcomings of traditional data applications such as weak intelligence and lagging transformation, and reshape the application form of intelligent information systems. Through keynote reports and roundtable discussions, the forum gathers industry experts and industrial elites to share the frontier technologies and latest industrial trends of Data Agents. It tackles technical difficulties and promotes innovation implementation through an industry-academia-research collaborative model, accelerates the large-scale innovative application of Data Agents, and empowers the digital-intelligent transformation and upgrading of industries in various fields.","is_demo":0,"cover_url":"/api/forums/4013/cover?size=thumb","cover_alt":"13th Frontier Forum on Intelligent Information Systems: Data Agents Empowering Industrial Digital-Intelligent Innovation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4013","source_checked_at":"2026-10-30","agenda_json":"[[\"Multimodal Agents Reshaping a New Paradigm of Full-Cycle Intelligent Healthcare\",\"张勇\"],[\"Building Trustworthy Data Agents from the OS Kernel\",\"田洪亮\"],[\"Large Models for Open Dynamic Environments: Knowledge Evolution, Cross-Domain Transfer, and Efficient Deployment\",\"尚学群\"],[\"Research and Application of the Autonomous and Controllable Zhipu GLM Large Model\",\"张鹏\"]]","code":"CNCC2026-4013","forum_type":"专题论坛","format":null,"theme_code":"Agent智能体","theme_name":"Agent智能体","venue":"西博城4号馆","room":"西博城4号馆-410","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4013","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第十三届智能信息系统前沿论坛：数据智能体赋能产业数智化创新","original_topic":"Agent智能体","original_description":"大模型技术持续驱动智能信息系统迭代升级，数据智能体（Data Agent）是新一代数据智能核心技术，可破解传统数据分析低效、价值释放不足痛点，是领域前沿热点。本次论坛聚焦Data Agent创新赋能，围绕核心技术与产业应用开展专题研讨。Data Agent依托大模型语义理解与智能编排，实现数据处理与决策全流程自主化，弥补了传统数据应用智能化弱、转化滞后的短板，重塑了智能信息系统应用形态。论坛通过主旨报告与圆桌讨论，汇聚业内专家与产业精英，分享Data Agent前沿技术与最新产业趋势。以产学研协同模式攻坚技术难点、推动创新落地，加速数据智能体规模化创新应用，赋能各领域产业数智化转型升级。","people":[{"forum_id":4013,"role":"chair","name":"邢春晓","affiliation":"清华大学","sort_order":0},{"forum_id":4013,"role":"guest","name":"张勇","affiliation":"清华大学","sort_order":0},{"forum_id":4013,"role":"guest","name":"田洪亮","affiliation":"蚂蚁技术研究院","sort_order":1},{"forum_id":4013,"role":"guest","name":"尚学群","affiliation":"西北工业大学","sort_order":2},{"forum_id":4013,"role":"guest","name":"张鹏","affiliation":"北京智谱华章科技有限公司","sort_order":3},{"forum_id":4013,"role":"guest","name":"王鑫","affiliation":"天津大学","sort_order":4}],"host":"","image_url":"/api/forums/4013/cover"},{"id":4014,"title":"Database and AI Technology Innovation","topic":"Data Science and Data Spaces","kind":"专题论坛","chair":"陈育兴、耿航","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-514","description":"Artificial intelligence technology is developing at an explosive and rapid pace, imposing higher requirements and new challenges on the storage, refined management, intelligent analysis, and value-based application of massive data, while also opening up broad room for growth. As the core foundation of全域 data management, databases' deep integration with artificial intelligence has become an industry trend. This sub-forum will bring together guests from universities, research institutes, and leading industry enterprises to conduct in-depth discussions on frontier innovative technologies and practical deployment scenarios around the two-way integration paths of DB For AI and AI For DB.","is_demo":0,"cover_url":"/api/forums/4014/cover?size=thumb","cover_alt":"Database and AI Technology Innovation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4014","source_checked_at":"2026-10-30","agenda_json":"[[\"Concluding Remarks\",\"耿航\"],[\"Opening Address\",\"刘婷婷\"],[\"A New Paradigm of Data Computing in the AI Era: From High-Performance Databases to Direct Computation on Compressed Data\",\"张峰\"],[\"Development of Traditional Relational Database Technology in the Era of Large Models\",\"陈璐\"],[\"Generative Databases: Current Status and Future\",\"张超\"],[\"Automated Data Analysis Based on Large Language Models\",\"朱君鹏\"],[\"From ATLAS to Nexa Knowledge: Research and Practice on Self-Evolving Semantic Layers for Large-Scale Databases\",\"张卿\"]]","code":"CNCC2026-4014","forum_type":"专题论坛","format":null,"theme_code":"数据科学与数据空间","theme_name":"数据科学与数据空间","venue":"西博城5号馆","room":"西博城5号馆-514","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":7,"forum_id":"CNCC2026-4014","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"数据库与AI技术创新","original_topic":"数据科学与数据空间","original_description":"人工智能技术呈爆发式高速发展态势，对海量数据的存储、精细化管理、智能分析及价值化应用，提出了更高要求与全新考验，也开辟了广阔增长空间。数据库作为全域数据管理的核心底座，与人工智能深度融合已是行业大势。本次分论坛将汇聚高校科研院所、行业龙头企业多方嘉宾，围绕 DB For AI、AI For DB 双向融合路径，深度研讨前沿创新技术与落地实践场景。","people":[{"forum_id":4014,"role":"chair","name":"陈育兴","affiliation":"腾讯科技（深圳）有限公司","sort_order":0},{"forum_id":4014,"role":"guest","name":"刘婷婷","affiliation":"腾讯科技（深圳）有限公司","sort_order":0},{"forum_id":4014,"role":"co_chair","name":"耿航","affiliation":"腾讯云计算（北京）有限责任公司","sort_order":1},{"forum_id":4014,"role":"guest","name":"张峰","affiliation":"中国人民大学","sort_order":1},{"forum_id":4014,"role":"guest","name":"陈璐","affiliation":"浙江大学","sort_order":2},{"forum_id":4014,"role":"guest","name":"张超","affiliation":"中国人民大学","sort_order":3},{"forum_id":4014,"role":"guest","name":"朱君鹏","affiliation":"西北农林科技大学","sort_order":4},{"forum_id":4014,"role":"guest","name":"张卿","affiliation":"腾讯云计算（北京）有限责任公司","sort_order":5}],"host":"","image_url":"/api/forums/4014/cover"},{"id":4016,"title":"AI Foresees 6G: Challenges, Technologies, and Applications","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"Tony Quek","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-403","description":"This forum focuses on \"AI-native 6G networks\" and explores the profound changes that artificial intelligence will bring to 6G architecture, protocols, resource management, security and trustworthiness, and business applications once it moves from an auxiliary tool to an intrinsic network capability. The forum will discuss the following questions: How will AI reshape 6G network architecture? How can the deep integration of communication, sensing, computing, and intelligence be achieved? What challenges do AI-native networks face in self-optimization, self-healing, low latency, green energy saving, and security protection? What new applications will they generate in scenarios such as industrial internet, low-altitude economy, intelligent transportation, and the metaverse? This topic aligns with the frontier trend of 6G and artificial intelligence integration, helps gather consensus among academia, industry, and application parties, promotes key technology breakthroughs, standards discussion, and ecosystem building, and provides directional reference for the future development of intelligent networks.","is_demo":0,"cover_url":"/api/forums/4016/cover?size=thumb","cover_alt":"AI Foresees 6G: Challenges, Technologies, and Applications","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4016","source_checked_at":"2026-10-30","agenda_json":"[[\"Task-Oriented Multi-Model Communication and Collaboration from an AI-RAN Perspective\",\"陈子晗\"],[\"Integrated Sensing, Communication, Computing, and Control Network Architecture and Closed-Loop Optimization for 6G Embodied Intelligence\",\"何欣欣\"],[\"LLMs for Next-Generation Communication Networks: Foundations, Key Technologies, and Future Directions\",\"Xue (Steve) Liu\"],[\"Speak Your Network: Automated Network Simulation Construction Based on Multi-Agent Orchestration\",\"虞红芳\"],[\"Key Technologies for Intelligent Endogenous Wireless Networks Empowered by Multi-Agent Collaboration\",\"夏文超\"]]","code":"CNCC2026-4016","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-403","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4016","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI预见6G：挑战、技术、与应用","original_topic":"新型计算与网络架构","original_description":"本论坛聚焦“AI原生的6G网络”，探讨人工智能从辅助工具走向网络内生能力后，对6G架构、协议、资源管理、安全可信与业务应用带来的深刻变革。论坛将围绕以下问题展开讨论：AI如何重塑6G网络架构？如何实现通信、感知、计算与智能的深度融合？AI原生网络在自优化、自愈合、低时延、绿色节能和安全防护方面面临哪些挑战？其在工业互联网、低空经济、智能交通、元宇宙等场景中将产生哪些新应用？该选题契合6G与人工智能融合发展的前沿趋势，有助于汇聚学术界、产业界和应用方共识，推动关键技术突破、标准研讨与生态建设，为未来智能网络发展提供方向参考。","people":[{"forum_id":4016,"role":"chair","name":"Tony Quek","affiliation":"新加坡科技设计大学","sort_order":0},{"forum_id":4016,"role":"guest","name":"陈子晗","affiliation":"新加坡科技设计大学","sort_order":0},{"forum_id":4016,"role":"guest","name":"Xue (Steve) Liu","affiliation":"MBZUAI","sort_order":1},{"forum_id":4016,"role":"guest","name":"虞红芳","affiliation":"电子科技大学","sort_order":2},{"forum_id":4016,"role":"guest","name":"何欣欣","affiliation":"北京邮电大学","sort_order":3},{"forum_id":4016,"role":"guest","name":"夏文超","affiliation":"南京邮电大学","sort_order":4}],"host":"","image_url":"/api/forums/4016/cover"},{"id":4021,"title":"Reconstructing the AI Compilation Stack Amid the Large Model Wave: Ascend NPU Practice and a New Paradigm for Domestic Software-Hardware Synergy","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"高耀清","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-110卧龙","description":"As the core bridge connecting software and hardware, compilation systems are undergoing profound changes brought about by AI technology. On the one hand, large language models are beginning to penetrate the compilation field, from assisting in generating optimization sequences to exploring the new end-to-end paradigm of \"AI as compiler\"; on the other hand, the explosive growth of AI workloads is forcing compilation systems to be reconstructed, and how to efficiently adapt to the endless stream of new hardware architectures has become a core issue. At the same time, the rise of domestic heterogeneous computing platforms imposes higher requirements on the agility and openness of the compilation stack.\nThis forum focuses on the construction and deployment of domestic AI compilation infrastructure, and provides an in-depth analysis of the key path to building a unified compilation architecture based on MLIR multi-level abstract intermediate representations. The forum will present a major sharing of the implementation experience of the Ascend NPU-IR project, the first in China to receive native Triton community certification, achieving efficient migration of AI applications onto domestic hardware and unleashing extreme performance.\nThis forum aims to bring together top experts from academia and industry to jointly promote the evolution of compilation systems toward intelligence, efficiency and openness, around directions such as the paradigm shift in compilation technology driven by AI, compilation optimization methods for AI workloads, and agile compilation frameworks for new hardware.","is_demo":0,"cover_url":"/api/forums/4021/cover?size=thumb","cover_alt":"Reconstructing the AI Compilation Stack Amid the Large Model Wave: Ascend NPU Practice and a New Paradigm for Domestic Software-Hardware Synergy","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4021","source_checked_at":"2026-10-30","agenda_json":"[[\"AscendNPU-IR: An Open-Source Compilation Foundation for Seamless Multi-Language Access to the Ascend Ecosystem\",\"海丽娟\"],[\"Full-Stack Optimization Methods for the Ascend AI Compiler: From Causal Decision-Making and Global Scheduling to Intelligent Kernel Generation\",\"赵捷\"],[\"AI Infrastructure in the Agent Era: Optimized Verifiability\",\"尚笠\"],[\"General-Computing Fusion Operator Generation Based on Ascend Multi-Core Task Orchestration\",\"李诚\"],[\"AI Compiler Optimization: Adaptation and Optimization of Triton-tle on Domestic Platforms\",\"柴赟达\"]]","code":"CNCC2026-4021","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"天府国际会议中心","room":"天府国际会议中心-110卧龙","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4021","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"大模型浪潮下的AI编译栈重构：昇腾NPU实践与国产软硬件协同新范式","original_topic":"新型计算与网络架构","original_description":"编译系统作为连接软硬件的核心桥梁，正经历着AI技术带来的深刻变革。一方面，大语言模型开始渗透编译领域，从辅助生成优化序列到探索端到端的“AI即编译器”新范式；另一方面，AI工作负载的爆发式增长也倒逼编译系统重构，如何高效适配层出不穷的新硬件架构成为核心议题。同时国产异构算力平台的崛起，对编译栈的敏捷性与开放性提出了更高要求。\n本论坛聚焦国产AI编译基础设施的建设与落地，深度解析基于MLIR多级抽象中间表示构建统一编译架构的关键路径。论坛将重磅分享国内首个获得原生Triton社区认证的Ascend NPU-IR项目的落地经验案例，实现AI应用在国产硬件上的高效迁移与极致性能释放。\n本论坛旨在汇聚学术界与工业界的顶尖专家，围绕AI驱动下的编译技术范式变革、面向AI工作负载的编译优化方法、以及面向新硬件的敏捷编译框架等方向，共同推动编译系统向智能化、高效化、开放化的方向演进。","people":[{"forum_id":4021,"role":"chair","name":"高耀清","affiliation":"华为公司","sort_order":0},{"forum_id":4021,"role":"guest","name":"海丽娟","affiliation":"华为","sort_order":0},{"forum_id":4021,"role":"guest","name":"赵捷","affiliation":"湖南大学","sort_order":1},{"forum_id":4021,"role":"guest","name":"尚笠","affiliation":"复旦大学","sort_order":2},{"forum_id":4021,"role":"guest","name":"李诚","affiliation":"中国科学技术大学","sort_order":3},{"forum_id":4021,"role":"guest","name":"柴赟达","affiliation":"先进编译实验室","sort_order":4},{"forum_id":4021,"role":"guest","name":"胡桥","affiliation":"华为技术有限公司","sort_order":5},{"forum_id":4021,"role":"guest","name":"刘威","affiliation":"华为技术有限公司","sort_order":6}],"host":"","image_url":"/api/forums/4021/cover"},{"id":4025,"title":"Frontier Agent Technologies: From the Digital World to the Physical World","topic":"Unclassified","kind":"专题论坛","chair":"张伟楠","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城2号馆-208","description":"Artificial intelligence is moving from digital intelligence interacting through virtual interfaces to embodied intelligence interacting with physical environments. Digital agents start from tool-based execution forms such as GUI agents, gradually evolve into personalized execution agents with long-term memory, personal adaptation, and experience accumulation capabilities, and further move toward autonomous execution agents with full-time memory, world cognition, and autonomous decision-making capabilities. Embodied agents, as the supporting entities of autonomous agents in the physical world, receive abstract decision instructions and complete perception, operation, and interaction in real space. This forum conducts discussions around key topics such as large-model-driven agent architectures, multimodal perception, planning and reasoning, memory mechanisms, GUI operation, embodied brains, and robot control. The forum aims to explore the technological trend of digital agents and embodied agents moving from independent evolution to cross-integration, analyze how the two empower each other bidirectionally, and jointly promote agents to achieve a capability leap from digital execution to physical action.","is_demo":0,"cover_url":"/api/forums/4025/cover?size=thumb","cover_alt":"Frontier Agent Technologies: From the Digital World to the Physical World","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4025","source_checked_at":"2026-10-30","agenda_json":"[[\"Toward Next-Generation General Agents: Personalization, Proactivity, and Sustainable Evolution\",\"刘卫文\"],[\"Harness-aware GUI Agents\",\"楼星雨\"],[\"Achieving General Embodied Intelligence Based on Hierarchical Architecture and Unified Protocols\",\"范鹤鹤\"],[\"From Stronger Models to Better Systems: Cost Reduction and Efficiency Enhancement for Agents and Long-Cycle Operation\",\"胡舒悦\"],[\"Open-World Generalizable Dexterous Manipulation\",\"许华哲\"]]","code":"CNCC2026-4025","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城2号馆","room":"西博城2号馆-208","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4025","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智能体前沿技术：从数字世界到物理世界","original_topic":"未分类","original_description":"人工智能正从虚拟界面交互的数字智能，迈向物理环境交互的具身智能。数字智能体以GUI智能体等工具化执行形态为起点，逐步演进为具备长效记忆、个人适配与经验沉淀能力的个人化执行智能体，并进一步迈向拥有全时记忆、世界认知和自主决策能力的自主化执行智能体。具身智能体则作为自主化智能体在物理世界中的支撑实体，承接抽象决策指令，完成真实空间中的感知、操作与交互。本论坛围绕大模型驱动的智能体架构、多模态感知、规划推理、记忆机制、GUI操作、具身大脑与机器人控制等关键议题展开研讨。论坛旨在探讨数字智能体与具身智能体从独立演进走向交叉融合的技术趋势，剖析二者如何双向赋能，共同推动智能体实现从数字执行到物理行动的能力跃迁。","people":[{"forum_id":4025,"role":"chair","name":"张伟楠","affiliation":"上海交通大学","sort_order":0},{"forum_id":4025,"role":"guest","name":"刘卫文","affiliation":"上海交通大学","sort_order":0},{"forum_id":4025,"role":"guest","name":"楼星雨","affiliation":"OPPO广东移动通信有限公司","sort_order":1},{"forum_id":4025,"role":"guest","name":"范鹤鹤","affiliation":"浙江大学","sort_order":2},{"forum_id":4025,"role":"guest","name":"胡舒悦","affiliation":"上海人工智能实验室","sort_order":3},{"forum_id":4025,"role":"guest","name":"许华哲","affiliation":"清华大学","sort_order":4},{"forum_id":4025,"role":"guest","name":"王俊","affiliation":"OPPO","sort_order":5}],"host":"","image_url":"/api/forums/4025/cover"},{"id":4029,"title":"AI Security and Privacy Protection","topic":"Unclassified","kind":"专题论坛","chair":"王小云","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-511","description":"Large models and AI systems are deeply penetrating highly sensitive business chains such as autonomous driving, financial transactions, and medical diagnosis. Their data-driven autonomous decision-making paradigm shifts the security focus from traditional perimeter protection to the security of AI itself. Against this backdrop, cryptography and trusted computing have been elevated from auxiliary hardening measures to core pillars for building AI security. This forum aims to bring together the AI security and cryptography communities to exchange views on threat models, key technical roadmaps, engineering deployment, and standardization, promote the formation of reusable security architectures and assessable security baselines, and accelerate the trusted deployment of AI in highly sensitive scenarios.","is_demo":0,"cover_url":"/api/forums/4029/cover?size=thumb","cover_alt":"AI Security and Privacy Protection","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4029","source_checked_at":"2026-10-30","agenda_json":"[[\"Security of Generative AI Models\",\"张勇东\"],[\"Data-Centric Agent Security\",\"胡侠\"],[\"AI Security: The Paradigm Evolution of Adversarial Learning\",\"杜天宇\"],[\"Outlook on AI-Empowered Resilient Space Cyber-Physical Security\",\"尤启迪\"],[\"Cryptanalysis of Pseudo-Random Error-Correcting Codes\",\"王安宇\"]]","code":"CNCC2026-4029","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-511","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4029","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"人工智能安全与隐私保护","original_topic":"未分类","original_description":"大模型与人工智能系统正深度渗透自动驾驶、金融交易、医疗诊断等高敏业务链路，其数据驱动的自主决策范式将安全重心从传统边界防护推向人工智能自身安全。在此背景下，密码学与可信计算从辅助加固手段，升格为构建人工智能安全的核心支柱。本论坛旨在汇聚人工智能安全与密码学社区，围绕威胁模型、关键技术路线、工程落地与标准化展开交流，推动形成可复用的安全架构与可评估的安全基线，加速人工智能在高敏场景中的可信部署。","people":[{"forum_id":4029,"role":"chair","name":"王小云","affiliation":"清华大学","sort_order":0},{"forum_id":4029,"role":"guest","name":"陈恺","affiliation":"中国科学院信息工程研究所","sort_order":0},{"forum_id":4029,"role":"guest","name":"王伟","affiliation":"西安交通大学","sort_order":1},{"forum_id":4029,"role":"guest","name":"李洪伟","affiliation":"电子科技大学","sort_order":2},{"forum_id":4029,"role":"guest","name":"孟国柱","affiliation":"中国科学院信息工程研究所","sort_order":3},{"forum_id":4029,"role":"guest","name":"张勇东","affiliation":"中国科学技术大学","sort_order":4},{"forum_id":4029,"role":"guest","name":"段海新","affiliation":"清华大学","sort_order":5},{"forum_id":4029,"role":"guest","name":"胡侠","affiliation":"上海人工智能实验室","sort_order":6},{"forum_id":4029,"role":"guest","name":"杜天宇","affiliation":"浙江大学","sort_order":7},{"forum_id":4029,"role":"guest","name":"尤启迪","affiliation":"北京卫星信息工程研究所","sort_order":8},{"forum_id":4029,"role":"guest","name":"王安宇","affiliation":"清华大学","sort_order":9}],"host":"","image_url":"/api/forums/4029/cover"},{"id":4040,"title":"Technical Development and Disciplinary Practice of AI-Native Research—When the Research Subject Shifts from Humans to Humans and AI","topic":"AI4Science","kind":"专题论坛","chair":"李贝","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-416","description":"Large models are upgrading from \"research tools\" to \"research collaborators,\" driving the research subject to evolve from a single person into a hybrid intelligence composed of humans and AI Agents. While this paradigm shift brings opportunities, it also poses new challenges to system architecture, human-computer interaction, and academic governance. This seminar focuses on the key technologies and ethical practices of AI-Native research, with topics covering: the design of Socratic guidance systems based on prompt engineering, the construction of RAG-driven research knowledge bases, the orchestration of automated research workflows with multi-Agent collaboration, hybrid scheduling architectures for cloud-based large-model inference and local GPU training, and Constitutional AI constraint mechanisms for academic integrity, inviting researchers and system developers to jointly explore a \"trustworthy, traceable, and inclusive\" AI-Native research ecosystem.\n\nForum content:\nThe topics cover research data infrastructure, accelerated computing, agent platforms, and disciplinary applications, and can serve as reference material for forum topic setting and speaker selection.\n\nWang Yuan, Head of the Product Management and Technical Architecture Department, Database Product Division, Alibaba Cloud Intelligence Group. Report title: \"Connecting Computing Power and Agents: A Data Foundation That Lets Research Agents Make Good Use of Every Piece of Data,\" discussing how research data can be efficiently stored, accessed, memorized, and retrieved by agents, as well as the ways in which data infrastructure connects with computing power systems.\n\nZhai Jian, Senior Solutions Architect at NVIDIA. Report title: \"NVIDIA Full-Stack Acceleration of Agentic AI and Physical AI,\" introducing the accelerated computing technology stack for agent applications and physical AI, covering the overall system from hardware to software frameworks.\n\nChen You, Head of Developer Growth Products at the Qwen AI Platform. Report title: \"Good Models, Good Platform: The Qwen AI Platform Is Born for AI Agents, Unleashing Research Productivity,\" introducing an agent platform that integrates models, tools, and runtime environments, and how researchers build, invoke, and run agents on it.\n\nWang Yuzhu, Director of the Department of Artificial Intelligence, School of Artificial Intelligence, China University of Geosciences (Beijing), Associate Professor, Doctoral Supervisor. Report title: \"Research on Key Technologies for Building Mineral Resource Large Models and Agents,\" introducing methods and progress of domain large models and agent construction technologies in mineral resource research, as representative work of large-model technology implementation in a specific discipline.\n\nJin Bowen, Assistant Professor at the School of History and the Center for Digital History, Nanjing University. Report title: \"AI and Historical Research,\" discussing the application of artificial intelligence methods in the organization, retrieval, and interpretation of historical materials and their impact on the ways of historical research, representing exploration in the direction of the humanities.","is_demo":0,"cover_url":"/api/forums/4040/cover?size=thumb","cover_alt":"Technical Development and Disciplinary Practice of AI-Native Research—When the Research Subject Shifts from Humans to Humans and AI","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4040","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4040","forum_type":"专题论坛","format":null,"theme_code":"AI4Science","theme_name":"AI4Science","venue":"西博城4号馆","room":"西博城4号馆-416","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":0,"forum_id":"CNCC2026-4040","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI原生科研的技术发展与学科实践——当研究主体从人变为人与AI","original_topic":"AI4Science","original_description":"大模型正在从\"科研工具\"升级为\"科研协作者\"，推动研究主体从单个人演变为人与AI Agent组成的混合智能体。这一范式跃迁在带来机遇的同时，也对系统架构、人机交互与学术治理提出新挑战。本研讨会聚焦AI-Native科研的关键技术与伦理实践，议题涵盖：基于提示词工程的苏格拉底式引导系统设计、RAG驱动的科研知识库构建、多Agent协作的自动化科研工作流编排、云端大模型推理与本地GPU训练的混合调度架构，以及面向学术诚信的Constitutional AI约束机制，邀请科研人员与系统开发者共探\"可信、可溯源、同普惠\"的AI-Native科研生态。\n\n论坛内容：\n议题覆盖科研数据基础设施、加速计算、智能体平台与学科应用，可作为论坛议题设置与讲者遴选的参考素材。\n\n王远，阿里云智能集团数据库产品事业部 产品管理与技术架构部负责人。报告题目《贯通算力与智能体：让科研智能体用好每一份数据的数据底座》，讨论科研数据如何被智能体高效存取、记忆与检索，以及数据基础设施与算力体系的衔接方式。\n\n翟健，NVIDIA 资深解决方案架构师。报告题目《NVIDIA 全栈加速 Agentic AI 和物理 AI》，介绍面向智能体应用与物理 AI 的加速计算技术栈，覆盖从硬件到软件框架的整体体系。\n\n辰悠，千问 AI 平台开发者增长产品负责人。报告题目《好模型，好平台：千问 AI 平台为 AI Agent 而生，释放科研生产力》，介绍将模型、工具与运行环境集成于一体的智能体平台，以及研究者在其上构建、调用并运行智能体的方式。\n\n王玉柱，中国地质大学（北京）人工智能学院人工智能系主任、副教授、博士生导师。报告题目《矿产资源大模型与智能体构建关键技术研究》，介绍领域大模型与智能体构建技术在矿产资源研究中的方法与进展，是大模型技术在具体学科落地的代表性工作。\n\n金伯文，南京大学历史学院暨数字史学研究中心助理教授。报告题目《AI 与历史学研究》，讨论人工智能方法在史料整理、检索与阐释中的应用及其对历史学研究方式的影响，代表人文学科方向的探索。","people":[{"forum_id":4040,"role":"chair","name":"李贝","affiliation":"阿里巴巴","sort_order":0}],"host":"","image_url":"/api/forums/4040/cover"},{"id":4045,"title":"Intelligent Computing System Ecosystem and Applications","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"陈全","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-409","description":"Demand for intelligent inference workloads has exploded, making them the most core and fastest-growing key workload. With the continuous deployment of scenarios such as autonomous driving and intelligent interaction, higher requirements are placed on the real-time performance, high concurrency, and scale of inference computing power. At present, the pace of platform architecture iteration, resource scheduling, and infrastructure evolution struggles to keep up with the development rhythm of inference workloads. The traditional general-purpose, static deployment and scheduling model of platforms cannot adapt to the characteristics of inference workloads—dynamic fluctuation, resource fragmentation, and stringent hardware adaptation requirements—resulting in a clear mismatch between the two in terms of time scale and operating mechanisms, which constrains computing power utilization efficiency and the large-scale deployment of the AI industry.\nTo this end, this forum gathers authoritative experts from industry, academia, and research to focus on the challenge of adapting inference workloads to computer systems, analyze key technical bottlenecks and development trends from the perspective of the full system stack, explore efficient adaptation solutions for hardware and intelligent workloads, and provide strong support for the high-quality development of China's artificial intelligence industry.","is_demo":0,"cover_url":"/api/forums/4045/cover?size=thumb","cover_alt":"Intelligent Computing System Ecosystem and Applications","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4045","source_checked_at":"2026-10-30","agenda_json":"[[\"Supercomputing in the AI Era: Acceleration and Integration\",\"卢宇彤\"],[\"Training and Inference Systems for Domestic Intelligent Computing Power\",\"翟季冬\"],[\"Memory Access and Communication Optimization for Supercomputing and Intelligent Computing\",\"董德尊\"],[\"Research on System Optimization Techniques for Large Models and Large-Model-Based Systems\",\"陈晨\"],[\"TokenOps: Scheduling and Optimization for Large Model Serving\",\"刘茼\"]]","code":"CNCC2026-4045","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-409","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4045","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智能计算系统生态及应用","original_topic":"新型计算与网络架构","original_description":"智能推理负载需求爆发式增长，已成为最核心、增长最快的关键负载。自动驾驶、智能交互等场景持续落地，对推理算力的实时性、高并发与规模化提出更高要求。现阶段，平台架构迭代、资源调度与基础设施演进速度，难以跟上推理负载的发展节奏。传统平台通用、静态的部署调度模式，无法适配推理负载动态波动、资源碎片化、硬件适配严苛的特征，导致二者在时间尺度与运行机制上出现明显失配，制约算力利用效率与AI产业规模化落地。\n为此，本次论坛汇聚产学研领域权威专家，聚焦推理负载与计算机系统适配难题，从全系统栈角度剖析关键技术瓶颈与发展趋势，探索硬件与智能负载的高效适配方案，为我国人工智能产业高质量发展提供有力支撑。","people":[{"forum_id":4045,"role":"chair","name":"陈全","affiliation":"上海交通大学","sort_order":0},{"forum_id":4045,"role":"guest","name":"卢宇彤","affiliation":"中山大学计算机学院 国家超算深圳/广州中心","sort_order":0},{"forum_id":4045,"role":"guest","name":"翟季冬","affiliation":"清华大学","sort_order":1},{"forum_id":4045,"role":"guest","name":"董德尊","affiliation":"国防科技大学","sort_order":2},{"forum_id":4045,"role":"guest","name":"陈晨","affiliation":"上海交通大学","sort_order":3},{"forum_id":4045,"role":"guest","name":"刘茼","affiliation":"联想（天津）有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4045/cover"},{"id":4050,"title":"\"Agent as a Service\" Empowers Service Industry Capacity Expansion and Quality Improvement: From Theoretical Breakthroughs to Industrial Practice","topic":"Unclassified","kind":"专题论坛","chair":"涂志莹","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城2号馆-205","description":"This forum closely follows General Secretary Xi Jinping's important instruction of \"demand traction, technology empowerment, and open cooperation,\" implements the State Council's strategic deployment for service industry capacity expansion and quality improvement, and conducts in-depth discussion of policies, theories, and technologies around \"consolidating the digital-intelligent foundation and empowering service industry capacity expansion and quality improvement\": analyzing the policy and strategic needs in the digital-intelligent service field and clarifying the disciplinary development positioning; discussing the new computing paradigm of \"Agent as a Service\" under the background of human-cyber-physical integration and building the underlying academic framework; exploring new application models driven by large models and agents and breaking through implementation bottlenecks in scenarios such as smart manufacturing, smart logistics, and technology finance.\nThe forum topic focuses on the urgent needs of high-quality development of the national service industry and the construction of a modern industrial system, aiming to consolidate the national digital-intelligent service foundation, cultivate an autonomous and controllable \"China Service\" technology brand, bridge the gap between frontier theories of agent service computing and large-scale industrial applications, and promote the transformation of the service industry from single-point digitalization to full-ecosystem digital-intelligent reconstruction.","is_demo":0,"cover_url":"/api/forums/4050/cover?size=thumb","cover_alt":"\"Agent as a Service\" Empowers Service Industry Capacity Expansion and Quality Improvement: From Theoretical Breakthroughs to Industrial Practice","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4050","source_checked_at":"2026-10-30","agenda_json":"[[\"\\\"Agent as a Service\\\" Empowering Service Industry Capacity Expansion and Quality Improvement: From Theoretical Breakthroughs to Industrial Practice (Forum Opening Introduction)\",\"涂志莹\"],[\"Agent-Based Metaverse Services and \\\"Agent as a Service (AaaS)\\\"\",\"徐晓飞\"],[\"Agent Services and Agent Service Networks\",\"尹建伟\"],[\"Agentic AI Infra System Software for Ubiquitous and Global Scenarios\",\"刘譞哲\"],[\"Agent Services for the Physical World: World Models, Executable Ontologies, and Heterogeneous Device Collaboration\",\"王牟江山\"],[\"Deterministic Agents for Local Life Services: Practical Exploration of Spatial Intent Understanding and LUI/GUI Interaction Paradigm Fusion\",\"李琳\"]]","code":"CNCC2026-4050","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城2号馆","room":"西博城2号馆-205","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4050","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"“智能体即服务”赋能服务业扩能提质：从理论突破到产业实践","original_topic":"未分类","original_description":"本论坛紧扣习近平总书记“需求牵引、科技赋能、开放合作”重要指示，落实国务院服务业扩能提质战略部署，围绕“夯实数智底座、赋能服务业扩能提质”开展政策、理论与技术的深度研讨：解析数智服务领域的政策战略需求，明确学科发展定位；研讨人机物融合背景下“智能体即服务”计算新范式，构建底层学术框架；探讨大模型与智能体驱动的应用新模式，破解智能制造、智慧物流、科技金融等场景的落地瓶颈。\n论坛选题聚焦国家服务业高质量发展与现代化产业体系建设的紧迫需求，旨在夯实国家数智服务底座，培育自主可控的“中国服务”科技品牌，填补智能体服务计算前沿理论与大规模产业应用之间的鸿沟，推动服务业从单点数字化向全生态数智化重构转型。","people":[{"forum_id":4050,"role":"chair","name":"涂志莹","affiliation":"哈尔滨工业大学（威海）","sort_order":0},{"forum_id":4050,"role":"guest","name":"徐晓飞","affiliation":"哈尔滨工业大学","sort_order":0},{"forum_id":4050,"role":"guest","name":"尹建伟","affiliation":"浙江大学","sort_order":1},{"forum_id":4050,"role":"guest","name":"刘譞哲","affiliation":"北京大学","sort_order":2},{"forum_id":4050,"role":"guest","name":"王牟江山","affiliation":"深圳开鸿数字产业发展有限公司","sort_order":3},{"forum_id":4050,"role":"guest","name":"李琳","affiliation":"北京美大智达科技有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4050/cover"},{"id":4055,"title":"AI High-Throughput Inference Infrastructure","topic":"Unclassified","kind":"专题论坛","chair":"李柳","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-118B峨眉厅B","description":"As large models move into the Agent era, Token has become the core value carrier of AI applications. With the theme \"Next-generation AI Infra for the Token Economy,\" this forum focuses on the requirements of high-throughput inference application scenarios, centering on the latest technological advances such as supernode PD-disaggregated architecture, DCU inference full-stack optimization, ParaCache advanced storage capability, and domestic IB-like high-speed networks. It comprehensively explores how to build a Token factory-level AI infrastructure with high throughput, low latency, controllable cost, and sustainable evolution through software-hardware collaborative innovation in computing, networking, and storage, accelerating AI's paradigm leap from answering questions to completing tasks.","is_demo":0,"cover_url":"/api/forums/4055/cover?size=thumb","cover_alt":"AI High-Throughput Inference Infrastructure","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4055","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Remarks\",\"魏振国\"],[\"Development Trends of Next-Generation Agent Infra for the Token Economy\",\"周震刚\"],[\"High-Throughput Inference in the Agent Era\",\"刘冠川\"],[\"FabricView: A High-Speed Network Operations and Management Platform\",\"赵欢\"],[\"AI Infra Software-Hardware Co-Design to Unleash Domestic Computing Power Efficiency\",\"师天麾\"],[\"Token-Driven Inference Efficiency Optimization for Large Models\",\"顾茹雅\"],[\"From Pilot Validation to Industrial Production: Practical Reflections on Mining Large Models\",\"樊启武\"],[\"Advancing Token-Efficient Throughput at the System Level\",\"孙建鹏\"],[\"Accelerating Token Generation with Advanced Storage Power\",\"石静\"],[\"Lightweight Reflections on AI Computing Infrastructure Development at Local Universities in the Agent Era\",\"陈超\"],[\"Supercomputing Internet: The Token Factory Network in the AI4S Era\",\"陶丹\"]]","code":"CNCC2026-4055","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-118B峨眉厅B","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":11,"forum_id":"CNCC2026-4055","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI 高通量推理基础设施","original_topic":"未分类","original_description":"随着大模型迈向 Agent 时代，Token 已成为AI应用的核心价值载体。本论坛以“面向 Token 经济的新一代 AI Infra”为主题，围绕高通量推理应用场景需求，聚焦超节点 PD 分离架构、DCU 推理全栈优化、ParaCache 先进存力与国产类 IB高速网络等最新技术进展，全方位探讨如何通过计算、网络、存储的软硬协同创新，构建高吞吐、低时延、成本可控、可持续演进的 Token 工厂级 AI 基础设施，加速 AI 从回答问题向完成任务的范式跃迁。","people":[{"forum_id":4055,"role":"chair","name":"李柳","affiliation":"曙光信息产业（北京）有限公司","sort_order":0},{"forum_id":4055,"role":"guest","name":"魏振国","affiliation":"曙光信息产业（北京）有限公司","sort_order":0},{"forum_id":4055,"role":"guest","name":"周震刚","affiliation":"IDC 咨询（北京）有限公司","sort_order":1},{"forum_id":4055,"role":"guest","name":"刘冠川","affiliation":"曙光信息产业股份有限公司","sort_order":2},{"forum_id":4055,"role":"guest","name":"赵欢","affiliation":"曙光信息产业（北京）有限公司","sort_order":3},{"forum_id":4055,"role":"guest","name":"师天麾","affiliation":"清程极智","sort_order":4},{"forum_id":4055,"role":"guest","name":"顾茹雅","affiliation":"基流科技","sort_order":5},{"forum_id":4055,"role":"guest","name":"樊启武","affiliation":"四川省自然资源数字科技有限责任公司","sort_order":6},{"forum_id":4055,"role":"guest","name":"孙建鹏","affiliation":"曙光智算","sort_order":7},{"forum_id":4055,"role":"guest","name":"石静","affiliation":"曙光信息产业（北京）有限公司","sort_order":8},{"forum_id":4055,"role":"guest","name":"陈超","affiliation":"四川轻化工大学","sort_order":9},{"forum_id":4055,"role":"guest","name":"陶丹","affiliation":"国家超算互联网","sort_order":10}],"host":"","image_url":"/api/forums/4055/cover"},{"id":4057,"title":"New-Architecture AI Chip Ecosystem Forum","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"欧阳鹏","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-407","description":"Aligned with the conference theme, this forum focuses on \"AI chip architecture innovation\" and \"domestic chip ecosystem development,\" bringing together core technical directions such as reconfigurable computing, wafer-scale integration, domain-specific architectures, and large-model adaptation and optimization, and is committed to providing a full-stack perspective and solutions from underlying hardware to upper-layer applications for the large-scale deployment of domestic new-architecture AI chips. The forum focuses on the full-stack closed loop of new-architecture AI chips—\"chip, computing, model, ecosystem\"—from chip architecture innovation to large-model adaptation and optimization, and from the full-stack AI software stack to developer ecosystem development, connecting the entire chain of industry, academia, research, and application.","is_demo":0,"cover_url":"/api/forums/4057/cover?size=thumb","cover_alt":"New-Architecture AI Chip Ecosystem Forum","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4057","source_checked_at":"2026-10-30","agenda_json":"[[\"Wafer-Scale Chips: A New Architectural Path to Break Through the AI Computing Power Dilemma\",\"胡杨\"],[\"Energy-Efficient Domain-Specific Architectures for Embodied Intelligence: Challenges and Practices\",\"贺光辉\"],[\"FlagOS: Making Every Chip Speak the Same Language\",\"林咏华\"],[\"End-to-End Full-Stack Optimization and Adaptation Technology for Reconfigurable Computing\",\"李彬\"],[\"From Models to Silicon: Chip Challenges and Closed-Loop Solutions from the Perspective of Large Model Vendors\",\"茹立云\"]]","code":"CNCC2026-4057","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-407","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4057","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"新架构AI芯片生态论坛","original_topic":"新型计算与网络架构","original_description":"本论坛紧扣大会主旨，聚焦「AI芯片架构创新」与「国产芯片生态建设」，汇聚可重构计算、晶圆级集成、领域专用架构、大模型适配优化等核心技术方向，致力于为国产新架构AI芯片的规模化落地提供从底层硬件到上层应用的全栈视角与解决方案。论坛聚焦新架构AI芯片「芯—算—模—生态」全栈闭环，从芯片架构创新到大模型适配优化，从全栈AI软件栈到开发者生态建设，打通产学研用全链条。","people":[{"forum_id":4057,"role":"chair","name":"欧阳鹏","affiliation":"北京清微智能科技股份有限公司","sort_order":0},{"forum_id":4057,"role":"guest","name":"胡杨","affiliation":"清华大学","sort_order":0},{"forum_id":4057,"role":"guest","name":"贺光辉","affiliation":"上海交通大学","sort_order":1},{"forum_id":4057,"role":"guest","name":"林咏华","affiliation":"北京智源人工智能研究院","sort_order":2},{"forum_id":4057,"role":"guest","name":"袁晟","affiliation":"北京智源人工智能研究院","sort_order":3},{"forum_id":4057,"role":"guest","name":"李彬","affiliation":"北京清微智能科技股份有限公司","sort_order":4},{"forum_id":4057,"role":"guest","name":"茹立云","affiliation":"本衍极智","sort_order":5}],"host":"","image_url":"/api/forums/4057/cover"},{"id":4059,"title":"Technical Evolution and Industrialization Path of On-Device Large Models for Phones: From Usability Validation to Product Deployment","topic":"AI Foundation Models","kind":"专题论坛","chair":"吴帆","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-505","description":"With the rapid development of large model technology, on-device AI is moving from proof of concept toward substantive deployment. This Workshop establishes \"on-device AI large models for phones have entered a deployable stage\" as its core assessment, aiming to systematically share frontline experience in the technical evolution and industrial application of on-device large models.\nThe conference will focus on the technical challenges currently faced by on-device AI, including on-device computing power constraints, the balance between accuracy and energy consumption, and privacy and security protection, and will systematically summarize the breakthrough progress the industry has made in the field of on-device deployment and optimization.\nThis Workshop will closely integrate OPPO's frontier exploration in the field of on-device AI, focusing on presenting OPPO's technical accumulation and practical achievements with academia during the deployment of on-device AI, hoping to provide the industry with a panoramic view from technology trend insights to engineering deployment, and jointly promote the maturation and evolution of on-device AI technology for phones.","is_demo":0,"cover_url":"/api/forums/4059/cover?size=thumb","cover_alt":"Technical Evolution and Industrialization Path of On-Device Large Models for Phones: From Usability Validation to Product Deployment","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4059","source_checked_at":"2026-10-30","agenda_json":"[[\"OPPO On-Device Computing in Practice\",\"宋晓辉\"],[\"On-Device AI Large Models Reshaping the Mobile Phone Industry Ecosystem: From Smart Terminals to a New Era of Agent Services\",\"魏亦豪\"],[\"Device-Cloud Collaborative Agent Systems\",\"郑臻哲\"],[\"On-Device AI Inference Acceleration: From Technical Breakthroughs to Product Deployment\",\"王锐\"],[\"From Smart Assistants to Personal Agents: The Evolution and Outlook of Future Mobile On-Device AI Models\",\"纪焘\"]]","code":"CNCC2026-4059","forum_type":"专题论坛","format":null,"theme_code":"AI基础模型","theme_name":"AI基础模型","venue":"西博城5号馆","room":"西博城5号馆-505","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4059","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"手机端侧大模型技术演进与产业化路径：从可用验证到产品落地","original_topic":"AI基础模型","original_description":"随着大模型技术的飞速发展，端侧化AI正从概念验证迈向实质落地阶段。本次 Workshop 确立“手机端侧AI大模型已进入可落地阶段”为核心研判，旨在系统分享端侧大模型在技术演进与产业应用中的一线经验。\n会议将重点研讨当前端侧AI面临的技术挑战，包括端侧算力约束、精度与能耗的均衡、隐私安全保护等，并系统总结行业在端侧部署与优化领域已取得的突破性进展。\n本次 Workshop 将紧密结合 OPPO 在端侧 AI 领域的前沿探索，重点呈现 OPPO 在端侧AI落地过程中与学术界的技术积累与实践成果，希望能为业界提供从技术趋势洞察到工程化落地的全景视图，共同推动手机端 AI 端侧化技术的成熟与演进。","people":[{"forum_id":4059,"role":"chair","name":"吴帆","affiliation":"上海交通大学","sort_order":0},{"forum_id":4059,"role":"guest","name":"宋晓辉","affiliation":"北京欧珀通信有限公司","sort_order":0},{"forum_id":4059,"role":"guest","name":"魏亦豪","affiliation":"亚马逊云科技","sort_order":1},{"forum_id":4059,"role":"guest","name":"郑臻哲","affiliation":"上海交通大学","sort_order":2},{"forum_id":4059,"role":"guest","name":"王锐","affiliation":"南京欧珀软件科技有限公司","sort_order":3},{"forum_id":4059,"role":"guest","name":"纪焘","affiliation":"复旦大学","sort_order":4}],"host":"","image_url":"/api/forums/4059/cover"},{"id":4062,"title":"Trustworthy Medical AI: Ethics, Governance and Technological Innovation","topic":"AI Safety and Governance","kind":"专题论坛","chair":"林俊宇","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-105都江堰","description":"Artificial intelligence, large models and other new-generation intelligent technologies are profoundly reshaping healthcare service models, bringing new development opportunities for disease diagnosis and treatment, clinical decision-making, health management and medical research, while also imposing higher requirements on ethical norms, risk prevention and control, regulatory governance and trustworthy applications. How to promote the safe, trustworthy, controllable and benevolent development of medical artificial intelligence has become an important issue of common concern to academia, industry and regulatory authorities. With the theme of \"Trustworthy Medical AI: Ethics, Governance and Technological Innovation,\" this forum invites well-known domestic experts in the field of medical AI ethics to engage in in-depth exchanges on hot topics such as the ethical governance system for medical artificial intelligence, ethical review mechanisms, ethical compliance risk assessment of large language models, and technology for good and human-machine symbiosis in the intelligent era, exploring new concepts, new methods and new practices for full-lifecycle governance of medical artificial intelligence, and providing theoretical support and practical paths for building a trustworthy medical AI ecosystem and promoting the high-quality development of medical artificial intelligence.","is_demo":0,"cover_url":"/api/forums/4062/cover?size=thumb","cover_alt":"Trustworthy Medical AI: Ethics, Governance and Technological Innovation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4062","source_checked_at":"2026-10-30","agenda_json":"[[\"Ethical and Governance Issues in Medical Artificial Intelligence\",\"雷瑞鹏\"],[\"Ethical Review of Medical Artificial Intelligence and Reflections on Its Boundaries\",\"陈蕾\"],[\"Accelerated Discovery and Degraded Capabilities: Ethical Reflections on Oncology Big Data Research in the AI Era\",\"张洪溢\"],[\"Risk Assessment of Medical Ethics and Compliance for Large Language Models\",\"徐汉辉\"],[\"Reflections on the Symbiosis of Technology and Humanity in the Intelligent Era\",\"李超\"]]","code":"CNCC2026-4062","forum_type":"专题论坛","format":null,"theme_code":"AI安全与治理","theme_name":"AI安全与治理","venue":"天府国际会议中心","room":"天府国际会议中心-105都江堰","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4062","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"可信医疗AI：伦理、治理与技术创新","original_topic":"AI安全与治理","original_description":"人工智能、大模型等新一代智能技术正深刻重塑医疗健康服务模式，为疾病诊疗、临床决策、健康管理和医学科研带来新的发展机遇，也对伦理规范、风险防控、监管治理和可信应用提出了更高要求。如何推动医疗人工智能实现安全、可信、可控、向善发展，已成为学术界、产业界和监管部门共同关注的重要议题。本论坛以“可信医疗AI：伦理、治理与技术创新”为主题，邀请国内医疗人工智能伦理领域知名专家，围绕医疗人工智能伦理治理体系、伦理审查机制、大语言模型伦理合规风险评估以及智能时代技术向善与人机共生等热点问题展开深入交流，探讨医疗人工智能全生命周期治理的新理念、新方法与新实践，为构建可信医疗AI生态、促进医疗人工智能高质量发展提供理论支撑和实践路径。","people":[{"forum_id":4062,"role":"chair","name":"林俊宇","affiliation":"复旦大学","sort_order":0},{"forum_id":4062,"role":"guest","name":"雷瑞鹏","affiliation":"电子科技大学","sort_order":0},{"forum_id":4062,"role":"guest","name":"陈蕾","affiliation":"四川大学华西临床医学院","sort_order":1},{"forum_id":4062,"role":"guest","name":"张洪溢","affiliation":"电子科技大学医学院","sort_order":2},{"forum_id":4062,"role":"guest","name":"徐汉辉","affiliation":"复旦大学","sort_order":3},{"forum_id":4062,"role":"guest","name":"李超","affiliation":"上海交大","sort_order":4}],"host":"","image_url":"/api/forums/4062/cover"},{"id":4068,"title":"The 5th Forum on Integrated Development of Open Source Ecosystem and Open Source Talent Cultivation","topic":"General","kind":"专题论坛","chair":"周傲英","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城2号馆-203","description":"Grounded in the national strategy of scientific and technological self-reliance, and directly addressing the shortage bottleneck of \"AI engineering talent,\" this forum focuses on the integrated development of the AI open source ecosystem and AI open source talent cultivation. We are committed to building a systematic ecosystem in which university teachers and students can continuously participate and gain growth and incentives. Centered on the evaluation and promotion system for AI open source contributions, we will discuss in depth tiered and categorized talent cultivation frameworks, knowledge systems, and scientific evaluation methods, and redefine a new paradigm for cultivating outstanding engineers in the AI era. The forum aims to help teachers and students succeed, empower industry, and serve national strategy. We sincerely invite leaders from universities, enterprises, and communities to jointly explore a new ecosystem of \"university-enterprise-community\" trinity collaborative education, promote education from \"knowledge and skills\" to \"real contributions,\" and provide solid support and action plans for the high-quality development of China's AI open source ecosystem.","is_demo":0,"cover_url":"/api/forums/4068/cover?size=thumb","cover_alt":"The 5th Forum on Integrated Development of Open Source Ecosystem and Open Source Talent Cultivation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4068","source_checked_at":"2026-10-30","agenda_json":"[[\"Scientifically Evaluating Open-Source Contributions and Reshaping the Growth Path of AI Engineers\",\"王伟\"],[\"AI-Driven Open Source Talent Development: Model Exploration and OpenTorial Practice\",\"游兰\"],[\"Real Projects, Real Work: Innovative Practices in Open Source Talent Development in the AI Era\",\"祝欣蓉\"],[\"From Real Projects to Sustained Contribution: Exploring Enterprise Open Source Talent Development in the AI Era\",\"董超\"],[\"Open Source Innovation Practices Based on Ascend Intelligent Computing Centers\",\"杨强\"],[\"Training Paths for Versatile International Communication Talent from the Perspective of the Global Open Source Movement: Practical Reflections Based on Wikipedia and the AI Open Source Ecosystem\",\"甘莅豪\"]]","code":"CNCC2026-4068","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城2号馆","room":"西博城2号馆-203","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4068","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第五届开源生态发展与开源人才培育一体化建设之路","original_topic":"综合","original_description":"立足国家科技自立自强战略，直面“AI 工程人才”短缺瓶颈，本论坛聚焦 AI 开源生态发展与 AI 开源人才培育的一体化建设。我们致力于构建一个能让高校师生持续参与、获得成长与激励的体系化生态，围绕人工智能开源贡献的评价与推动体系，深入探讨分层分类的人才培养框架、知识体系与科学评价方法，重新定义 AI 时代卓越工程师培育的新范式。论坛旨在成就师生、赋能产业、服务国家战略，诚邀高校、企业、社区引领者，共探“高校-企业-社区”三位一体协同育人新生态，推动教育从“知识技能”迈向“真实贡献”，为我国人工智能开源生态的高质量发展提供坚实支撑与行动方案。","people":[{"forum_id":4068,"role":"chair","name":"周傲英","affiliation":"华东师范大学","sort_order":0},{"forum_id":4068,"role":"guest","name":"王伟","affiliation":"华东师范大学","sort_order":0},{"forum_id":4068,"role":"guest","name":"游兰","affiliation":"湖北大学","sort_order":1},{"forum_id":4068,"role":"guest","name":"祝欣蓉","affiliation":"上海杉达学院","sort_order":2},{"forum_id":4068,"role":"guest","name":"董超","affiliation":"蚂蚁集团","sort_order":3},{"forum_id":4068,"role":"guest","name":"杨强","affiliation":"智云数科卓越研究院","sort_order":4},{"forum_id":4068,"role":"guest","name":"甘莅豪","affiliation":"华东师范大学","sort_order":5}],"host":"","image_url":"/api/forums/4068/cover"},{"id":4073,"title":"Intelligent Imaging for Smart Healthcare: Data Governance, Algorithm Breakthroughs, and Clinical Translation in Medical Imaging","topic":"Unclassified","kind":"专题论坛","chair":"胡伏原、丁熠","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-412","description":"This forum focuses on frontier directions such as medical imaging data governance, intelligent reconstruction, multimodal fusion, trustworthy artificial intelligence, and the intelligentization of medical equipment, and explores how artificial intelligence technologies can drive the transformation of medical imaging from data resources into clinical value. It will also discuss the following questions: (1) How can medical imaging AI improve model interpretability, clinical trustworthiness, and structural fidelity in reconstruction—how can black-box models win doctors' trust? (2) How can the gaps in multicenter medical imaging data be bridged—who sets the rules of the game? (3) How can the \"independent and controllable\" ecosystem and technical path for high-end medical equipment be explored? The five speakers at this forum cover the complete innovation chain of \"algorithm → equipment → clinical → regional application,\" closely aligning with the CNCC2026 theme \"Data-Driven Intelligence, Computing the Future,\" responding to the national \"AI + healthcare\" strategy, with discussion questions that directly address the core of medical-engineering cross-disciplinary translation, combining academic frontier relevance with clinical applicability.","is_demo":0,"cover_url":"/api/forums/4073/cover?size=thumb","cover_alt":"Intelligent Imaging for Smart Healthcare: Data Governance, Algorithm Breakthroughs, and Clinical Translation in Medical Imaging","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4073","source_checked_at":"2026-10-30","agenda_json":"[[\"LLM-Driven Medical Image Analysis\",\"夏勇\"],[\"Research on Instance Segmentation Methods for Head and Neck Lymph Node MRI Images\",\"周涛\"],[\"Team Practice from Medical Imaging and Digital Pathology to Biomarkers and Large Language Models\",\"庞明辉\"],[\"The New Era of Compute Tokenization: Practical Deployment of the \\\"Token Study Room\\\" and \\\"Token Campus Network\\\" to Make Research Agents Available to Everyone\",\"冀文辉\"],[\"Key Technology Breakthroughs in Domestic Computing Power for Intelligent Medical Image Computing\",\"肖贵宝\"],[\"New Advances in AI-Driven Fetal Imaging\",\"郭应坤\"]]","code":"CNCC2026-4073","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-412","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4073","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智影慧医：智能医学影像数据治理、算法突破与临床转化","original_topic":"未分类","original_description":"本论坛聚焦医学影像数据治理、智能重建、多模态融合、可信人工智能以及医疗装备智能化等前沿方向，探讨人工智能技术如何推动医学影像从数据资源向临床价值转化。并研讨以下问题：（1）医学影像AI如何提升模型可解释性、临床可信度以及重建结构保真度——黑箱模型如何赢得医生信赖？（2）多中心医学影像数据的差距如何弥补——谁来制定游戏规则？（3）高端医疗装备的\"自主可控\"生态与技术路径如何探索？本论坛五位嘉宾覆盖\"算法→装备→临床→区域应用\"完整创新链条，紧扣CNCC2026\"数聚驱动，智算未来\"主题，响应国家\"AI+医疗\"战略，思辨问题直扣医工交叉转化核心，兼具学术前沿性与临床落地性。","people":[{"forum_id":4073,"role":"chair","name":"胡伏原","affiliation":"苏州科技大学","sort_order":0},{"forum_id":4073,"role":"guest","name":"夏勇","affiliation":"西北工业大学","sort_order":0},{"forum_id":4073,"role":"co_chair","name":"丁熠","affiliation":"电子科技大学","sort_order":1},{"forum_id":4073,"role":"guest","name":"周涛","affiliation":"北方民族大学","sort_order":1},{"forum_id":4073,"role":"guest","name":"庞明辉","affiliation":"四川省人民医院","sort_order":2},{"forum_id":4073,"role":"guest","name":"冀文辉","affiliation":"联泰集群","sort_order":3},{"forum_id":4073,"role":"guest","name":"肖贵宝","affiliation":"中科曙光","sort_order":4},{"forum_id":4073,"role":"guest","name":"郭应坤","affiliation":"四川大学华西第二医院","sort_order":5},{"forum_id":4073,"role":"guest","name":"徐峰磊","affiliation":"苏州科技大学","sort_order":6}],"host":"","image_url":"/api/forums/4073/cover"},{"id":4093,"title":"High-Reliability Collaborative Low-Altitude Intelligence Technology and Applications in Hilly and Mountainous Areas","topic":"AI for Industry","kind":"专题论坛","chair":"牛宪华、陈晓亮","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-411","description":"This forum focuses on the high-reliability collaboration and safety issues of low-altitude intelligent systems in the complex environments of hilly and mountainous areas. Top experts in the fields of multi-agent systems, UAV control, distributed computing, and network and information security are invited to conduct in-depth discussions on core technologies such as resilient networking in obstructed scenarios, distributed collaborative decision-making, multi-UAV collaborative collision avoidance, communication anti-interference and attack detection, fault tolerance, and safety isolation, and to explore technical paths for highly reliable and safe collaboration under communication-constrained and strongly interfered environments, providing theoretical and technical support for typical low-altitude applications in hilly and mountainous areas such as low-altitude logistics, agricultural and forestry operations, emergency rescue, mountain inspection, and low-altitude traffic management.","is_demo":0,"cover_url":"/api/forums/4093/cover?size=thumb","cover_alt":"High-Reliability Collaborative Low-Altitude Intelligence Technology and Applications in Hilly and Mountainous Areas","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4093","source_checked_at":"2026-10-30","agenda_json":"[[\"Integrated Security Considerations for Complex Low-Altitude Scenarios\",\"伊鹏\"],[\"Intelligent Security Protection Technology and Applications for UAVs\",\"沈玉龙\"],[\"Heterogeneous Resource Collaboration Technologies for Low-Altitude Intelligent Networks\",\"冷甦鹏\"],[\"Air-Ground Cross-Domain Autonomous Collaboration in Practice: From Multirotors to Helicopters\",\"张卫东\"],[\"Safety Testing of the Intended Functionality of Unmanned Intelligent Systems\",\"张源\"]]","code":"CNCC2026-4093","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"西博城4号馆","room":"西博城4号馆-411","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4093","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"丘陵山区低空智能高可靠协同技术及其应用","original_topic":"AI+行业","original_description":"本论坛聚焦丘陵山区复杂环境下低空智能系统的高可靠协同与安全问题，邀请多智能体系统、无人机控制、分布式计算、网络与信息安全领域顶尖专家，围绕遮挡场景弹性组网、分布式协同决策、多机协同防碰撞、通信抗干扰与攻击检测、故障容错与安全隔离等核心技术展开深度研讨，探索通信受限、强干扰环境下高可靠安全协同的技术路径，为低空物流、农林作业、应急救援、山地巡检、低空交通管理等丘陵山区典型低空应用提供理论与技术支撑。","people":[{"forum_id":4093,"role":"chair","name":"牛宪华","affiliation":"西华大学","sort_order":0},{"forum_id":4093,"role":"guest","name":"伊鹏","affiliation":"嵩山实验室","sort_order":0},{"forum_id":4093,"role":"co_chair","name":"陈晓亮","affiliation":"西华大学","sort_order":1},{"forum_id":4093,"role":"guest","name":"沈玉龙","affiliation":"西安电子科技大学","sort_order":1},{"forum_id":4093,"role":"guest","name":"张卫东","affiliation":"上海交通大学","sort_order":2},{"forum_id":4093,"role":"guest","name":"冷甦鹏","affiliation":"电子科技大学","sort_order":3},{"forum_id":4093,"role":"guest","name":"张源","affiliation":"复旦大学","sort_order":4}],"host":"","image_url":"/api/forums/4093/cover"},{"id":4094,"title":"The Next Architectural Choice for AI-Era Data Center Network Foundations","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"祁楠","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-401","description":"Full title: Fast-Narrow Or Slow-Wide: The Next Architectural Choice for AI-Era Data Center Network Foundations\n\nLarge AI models are driving the rapid evolution of data center networks, and GPU cluster scale continues to expand. Interconnect has become a key factor determining the performance, energy consumption, cost, and scalability of AI infrastructure. Around the Fast-Narrow route represented by 112G/224G/448G rate increases, and the Slow-Wide route characterized by parallel lane expansion, the industry is facing a new architectural choice. This seminar will invite experts from data center operators, GPU/AI chips, switch chips, PHY, copper cables, optical interconnect, and new types of optical fiber, and will use a combination of keynote talks and roundtable discussions to explore the technical paths and industry collaboration for future AI data center network foundations, and to promote the formation of forward-looking consensus.","is_demo":0,"cover_url":"/api/forums/4094/cover?size=thumb","cover_alt":"The Next Architectural Choice for AI-Era Data Center Network Foundations","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4094","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Introduction\",\"祁楠\"],[\"Outlook on 448G/lane Generation High-Speed Interconnects\",\"何向\"],[\"Research on Key Technologies for Analog Mixed-Signal High-Speed Data Interfaces\",\"赵潇腾\"],[\"The Landscape and Choices of AI Compute Interconnects in the 448G Era\",\"吴国继\"],[\"Enhancing GPU Memory Pool Capacity and Bandwidth with Optical I/O\",\"韩佳巍\"],[\"Reflections on I/O Design in the AI Era\",\"刘西柯\"],[\"Optoelectronic Integrated High-Density Optical Interconnects for AI Computing\",\"肖钰\"],[\"Progress in Micro-LED-Based Parallel Matrix Optical Interconnect Technology\",\"杜源\"],[\"Glass Waveguide Substrates and Interface Technologies for Co-Packaged Optics\",\"马麟\"],[\"Fireside chat: Architectural Innovation in Slow-Wide vs Fast-Narrow Interfaces and Interconnect Technologies\",\"米光灿\"]]","code":"CNCC2026-4094","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-401","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":10,"forum_id":"CNCC2026-4094","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI时代数据中心网络底座的下一次架构抉择","original_topic":"新型计算与网络架构","original_description":"完整标题：Fast-Narrow Or Slow-Wide：AI时代数据中心网络底座的下一次架构抉择\n\n人工智能大模型推动数据中心网络快速演进，GPU集群规模持续扩大，互连已成为决定AI基础设施性能、能耗、成本与可扩展性的关键因素。围绕112G/224G/448G速率提升所代表的Fast-Narrow路线，以及以并行通道扩展为特征的Slow-Wide路线，产业界正面临新的架构选择。本研讨会将邀请来自数据中心运营商、GPU/AI芯片、交换芯片、PHY、铜缆、光互连与新型光纤等领域的专家，采用主题报告与圆桌交流相结合的形式，探讨未来AI数据中心网络底座的技术路径与产业协同，促进形成前瞻性共识。","people":[{"forum_id":4094,"role":"chair","name":"祁楠","affiliation":"中国科学院半导体研究所","sort_order":0},{"forum_id":4094,"role":"guest","name":"何向","affiliation":"北京华为数字技术有限公司","sort_order":0},{"forum_id":4094,"role":"guest","name":"赵潇腾","affiliation":"西安电子科技大学","sort_order":1},{"forum_id":4094,"role":"guest","name":"吴国继","affiliation":"安费诺通信方案事业部","sort_order":2},{"forum_id":4094,"role":"guest","name":"韩佳巍","affiliation":"沐曦科技（北京）有限公司","sort_order":3},{"forum_id":4094,"role":"guest","name":"刘西柯","affiliation":"默升科技（上海）有限公司","sort_order":4},{"forum_id":4094,"role":"guest","name":"肖钰","affiliation":"上海澜昆微电子科技有限公司","sort_order":5},{"forum_id":4094,"role":"guest","name":"杜源","affiliation":"南京大学","sort_order":6},{"forum_id":4094,"role":"guest","name":"马麟","affiliation":"上海交通大学","sort_order":7},{"forum_id":4094,"role":"guest","name":"米光灿","affiliation":"华为技术有限公司","sort_order":8}],"host":"","image_url":"/api/forums/4094/cover"},{"id":4096,"title":"The 9th Smart Cities and Urban Computing Forum—Spatiotemporal Intelligence and Urban Computing","topic":"AI for Industry","kind":"专题论坛","chair":"张钧波","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-414","description":"After years of development, smart cities have achieved phased results, formed rich application scenarios, and accumulated massive amounts of urban data. At the same time, the new generation of artificial intelligence technologies represented by large models, spatiotemporal AI, and embodied intelligence has brought new ideas and new opportunities for the continued development of smart cities.\n\nThis forum will invite leading figures from academia and industry in the fields of urban computing, artificial intelligence, and big data to focus on the real needs and major challenges of the physical world of cities, and to conduct in-depth exploration of spatiotemporal intelligence and physical AI around the theoretical foundations, technological frontiers, and application scenarios of urban computing, and, combining academic research and industrial practice, share the latest application implementation cases, leading smart cities toward a more intelligent and sustainable future.","is_demo":0,"cover_url":"/api/forums/4096/cover?size=thumb","cover_alt":"The 9th Smart Cities and Urban Computing Forum—Spatiotemporal Intelligence and Urban Computing","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4096","source_checked_at":"2026-10-30","agenda_json":"[[\"Spatiotemporal AI: Foundational Theories and Key Technologies for AI Entering the Physical World\",\"郑宇\"],[\"Applications of Digital Technology in Urban Infectious Disease Prevention and Control\",\"杨鹏\"],[\"Sensing-Computing Collaborative Intelligent IoT Systems\",\"郭斌\"],[\"World Models: From Cities to Embodiment\",\"李勇\"],[\"Interactive Data Intelligence Integration and Analysis\",\"巫英才\"]]","code":"CNCC2026-4096","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"西博城4号馆","room":"西博城4号馆-414","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4096","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第九届智慧城市与城市计算论坛——时空智能与城市计算","original_topic":"AI+行业","original_description":"经过多年建设，智慧城市已取得阶段性成果，形成了丰富的应用场景，沉淀了海量的城市数据。与此同时，以大模型、时空AI、具身智能为代表的新一代人工智能技术，为智慧城市的持续发展带来了新思路与新机遇。\n\n本论坛将邀请城市计算、人工智能、大数据领域的学术界与产业界领军人物，聚焦城市物理世界的真实需求与重大挑战，围绕城市计算的理论基础、技术前沿与应用场景，深入探究时空智能与物理AI，并结合学术研究与产业实践，分享最新应用落地案例，引领智慧城市迈向更加智能化与可持续化的未来。","people":[{"forum_id":4096,"role":"chair","name":"张钧波","affiliation":"京东科技控股股份有限公司","sort_order":0},{"forum_id":4096,"role":"guest","name":"郑宇","affiliation":"京东集团","sort_order":0},{"forum_id":4096,"role":"guest","name":"杨鹏","affiliation":"北京市疾病预防控制中心","sort_order":1},{"forum_id":4096,"role":"guest","name":"郭斌","affiliation":"西北工业大学","sort_order":2},{"forum_id":4096,"role":"guest","name":"李勇","affiliation":"清华大学","sort_order":3},{"forum_id":4096,"role":"guest","name":"巫英才","affiliation":"浙江大学","sort_order":4}],"host":"","image_url":"/api/forums/4096/cover"},{"id":4105,"title":"Efficient Neural Network Methods","topic":"Machine Learning Theory and Mechanisms","kind":"专题论坛","chair":"吕建成","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城2号馆-202","description":"Currently, artificial intelligence is shifting from \"parameter expansion\" to a critical stage of \"efficiency first\"—although large models have excellent capabilities, their hundreds of billions of parameters and extremely high computational consumption make it difficult to truly deploy them in resource-constrained real scenarios such as industrial production, autonomous driving, and mobile terminals. How to break the inherent paradigm of \"compute equals performance\" and enable AI to remain efficient, stable, and secure under milliwatt-level power consumption, hundred-megabyte-level memory, and dynamic open environments has become a core proposition urgently needing breakthrough in both academia and industry. This forum focuses on frontier directions such as efficient neural network architecture design and learning algorithms, promoting artificial intelligence from laboratory accuracy to broad and reliable real-world application.","is_demo":0,"cover_url":"/api/forums/4105/cover?size=thumb","cover_alt":"Efficient Neural Network Methods","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4105","source_checked_at":"2026-10-30","agenda_json":"[[\"Efficient Spiking Transformer Architecture and Computing Methods\",\"田永鸿\"],[\"Efficient Attention Mechanisms for Long-Sequence Modeling\",\"黄高\"],[\"Understanding Transformers from an Energy Perspective\",\"刘勇\"],[\"Dynamic Synaptic Spatiotemporal Information Processing Mechanisms and Advantages\",\"于强\"],[\"Vision-Centric Decentralized Multimodal Neural Computing Architecture: Neural Community Representation, Communication, and Collaboration\",\"王树徽\"],[\"Brain-Inspired Large Models and Their Applications\",\"李国齐\"]]","code":"CNCC2026-4105","forum_type":"专题论坛","format":null,"theme_code":"机器学习理论与机理","theme_name":"机器学习理论与机理","venue":"西博城2号馆","room":"西博城2号馆-202","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4105","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"高效神经网络方法","original_topic":"机器学习理论与机理","original_description":"当前，人工智能正从“参数膨胀”转向“效能优先”的关键阶段——大模型虽能力卓越，但千亿级参数量和极高算力消耗使其难以真正部署于工业生产、自动驾驶、移动终端等资源受限的真实场景。如何打破“算力即性能”的固有范式，让AI在毫瓦级功耗、百兆级内存和动态开放环境中依然保持高效、稳定与安全，已成为学术界与产业界亟待突破的核心命题。本论坛聚焦高效神经网络架构设计、学习算法等前沿方向，推动人工智能从实验室精度走向广泛而可靠的实战应用。","people":[{"forum_id":4105,"role":"guest","name":"田永鸿","affiliation":"北京大学","sort_order":0},{"forum_id":4105,"role":"chair","name":"吕建成","affiliation":"四川大学","sort_order":0},{"forum_id":4105,"role":"guest","name":"黄高","affiliation":"清华大学","sort_order":1},{"forum_id":4105,"role":"guest","name":"刘勇","affiliation":"中国人民大学","sort_order":2},{"forum_id":4105,"role":"guest","name":"于强","affiliation":"天津大学","sort_order":3},{"forum_id":4105,"role":"guest","name":"王树徽","affiliation":"中国科学院计算技术研究所","sort_order":4},{"forum_id":4105,"role":"guest","name":"李国齐","affiliation":"中国科学院自动化研究所","sort_order":5}],"host":"","image_url":"/api/forums/4105/cover"},{"id":4110,"title":"CTO Summit: Intelligence and Evolution: Co-evolution of Technology Paradigms and Organizations (Closed-door)","topic":"General","kind":"专题论坛","chair":"李大海","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-114瓦屋山","description":"The CTO Summit continues its small-scale, targeted closed-door format, setting two main threads: technology development and organizational evolution. Centered on the core theme \"CTO foresight, choices, and adaptation in the AI era,\" it builds a complete discussion logic:\nTechnology dimension: focusing on real-world issues such as the implementation of large-model AI Agents, security governance of world models, bottlenecks in the development of domestic computing power, and decision-making on large-scale computing power investment, directly addressing frontline challenges such as enterprise computing power procurement, AI transformation of existing systems, and loss of technical assets.\nOrganization dimension: focusing on frontier propositions such as building AI-native organizations, performance and responsibility division for digital employees, human-machine collaborative work models, the impact of super-individuals on enterprise architecture, and shifts in the focus of future organizational management, exploring the iteration path of enterprise management systems under the AI wave.\nAfter the meeting, the \"2026-2027 CTO Trend Industry Report\" will be compiled and produced, with the names of all participating contributors fully credited.","is_demo":0,"cover_url":"/api/forums/4110/cover?size=thumb","cover_alt":"CTO Summit: Intelligence and Evolution: Co-evolution of Technology Paradigms and Organizations (Closed-door)","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4110","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4110","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"天府国际会议中心","room":"天府国际会议中心-114瓦屋山","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":0,"forum_id":"CNCC2026-4110","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"CTO Summit 智能与演进：技术范式与组织的协同进化（闭门）","original_topic":"综合","original_description":"CTO Summit延续小规模定向闭门机制，设置技术发展与组织进化两大主线，围绕\"AI时代CTO的预判、选择与应变\"核心主题，搭建完整讨论逻辑：\n技术维度：重点围绕大模型Agent落地、世界模型安全管控、国产算力发展瓶颈、大规模算力投入决策等现实问题，直面企业算力采购、存量系统AI改造、技术资产流失等一线难题。\n组织维度：聚焦AI原生组织搭建、数字员工绩效与权责划分、人机协同工作模式、超级个体对企业架构的冲击、未来组织管理重心变迁等前沿命题，探讨AI浪潮下企业管理体系迭代路径。\n会后汇总编制《2026-2027年度CTO趋势行业报告》，完整标注全部参会贡献者姓名。","people":[{"forum_id":4110,"role":"chair","name":"李大海","affiliation":"面壁智能","sort_order":0},{"forum_id":4110,"role":"guest","name":"王巍","affiliation":"新浪微博","sort_order":0},{"forum_id":4110,"role":"guest","name":"王海峰","affiliation":"百度在线网络技术（北京）有限公司","sort_order":1},{"forum_id":4110,"role":"guest","name":"李斌","affiliation":"曙光信息产业（北京）有限公司","sort_order":2},{"forum_id":4110,"role":"guest","name":"韦青","affiliation":"微软","sort_order":3},{"forum_id":4110,"role":"guest","name":"陈刚","affiliation":"烽火通信科技股份有限公司战略研究院","sort_order":4},{"forum_id":4110,"role":"guest","name":"方国伟","affiliation":"华为技术有限公司","sort_order":5},{"forum_id":4110,"role":"guest","name":"薄满辉","affiliation":"中航信移动科技有限公司","sort_order":6}],"host":"","image_url":"/api/forums/4110/cover"},{"id":4111,"title":"Second Structural Pattern Recognition and Graph Machine Learning: Can Large Models and Graph Learning Meet Halfway?","topic":"AI Foundation Models","kind":"专题论坛","chair":"白璐","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-103剑门关","description":"Structural pattern recognition and graph machine learning are the primary methods in artificial intelligence for analyzing graph data or complex networks. They can effectively mine and learn the complex structural pattern information contained within, have been widely applied in many fields, and represent an important research direction of the new generation of artificial intelligence. However, current research still faces a series of key challenges and bottlenecks, such as over-smoothing and over-compression in graph representation learning, and model interpretability. In addition, the impact of large models in recent years is making Graph+LLM a frontier focus in this field. To promote in-depth exchange among scholars in this field, analyze current theoretical bottlenecks, and thereby propose new insights and new reflections that drive the field's continued development, this forum invites several well-known leading and outstanding scholars in this field in China to share their latest achievements in theoretical research and frontier interdisciplinary work. They will discuss frontier issues such as how large language models can mine graph structural priors to enhance generation and reasoning capabilities, how graph models can provide structured indexing and interpretable reasoning chains for LLMs, how large language models can empower the development of structural pattern recognition, and unified architectures for multimodal graph foundation models. This forum is applied for under the guidance and support of the CCF-AI Graph Machine Learning Group and is an important part of the group's annual academic activities.","is_demo":0,"cover_url":"/api/forums/4111/cover?size=thumb","cover_alt":"Second Structural Pattern Recognition and Graph Machine Learning: Can Large Models and Graph Learning Meet Halfway?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4111","source_checked_at":"2026-10-30","agenda_json":"[[\"Toward General Large Models for Structured Data\",\"崔鹏\"],[\"Understanding the Structured Language of Computation, Nature, and Society\",\"魏哲巍\"],[\"Graph Kernel Neural Networks: Exploration and Practice in Bringing Traditional Structural Pattern Similarity Information Back to Deep Learning\",\"白璐\"],[\"Graph-Structured Data Mining with Large-Model Agents: From Semantic Matching to Autonomous Interaction\",\"刘淇\"],[\"Deep Graph Sparse Learning and Robust Graph Pattern Recognition\",\"江波\"]]","code":"CNCC2026-4111","forum_type":"专题论坛","format":null,"theme_code":"AI基础模型","theme_name":"AI基础模型","venue":"天府国际会议中心","room":"天府国际会议中心-103剑门关","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4111","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第二届结构模式识别与图机器学习：大模型与图学习能否双向奔赴","original_topic":"AI基础模型","original_description":"结构模式识别与图机器学习是人工智能领域分析图数据或复杂网络的主要方法，能够有效挖掘与学习其内部蕴含的复杂结构模式信息，已在诸多领域获得广泛应用，是新一代人工智能重要研究方向。然而，当前研究依然面临一系列关键挑战与瓶颈，如：图表示学习的过平滑与过度压缩、模型的可解释性等。此外，近年大模型的冲击，正使得Graph+LLM成为该领域前沿焦点。为促进该领域学者深入交流，剖析当前理论瓶颈问题，进而提出推动该领域持续发展的新见解与新思考，本论坛邀请数位国内该领域一线知名领军与拔尖学者，分享其在理论研究与前沿交叉等方面的最新成果，并将讨论大语言模型如何挖掘图结构先验增强生成与推理能力、图模型怎样为LLM提供结构化索引与可解释推理链、大语言模型如何赋能结构模式识别领域的发展，以及多模态图基础模型的统一架构等前沿问题。本论坛是在CCF-AI 图机器学习学组的指导与支持下申办的，是学组年度学术活动的重要组成部分。","people":[{"forum_id":4111,"role":"guest","name":"崔鹏","affiliation":"清华大学","sort_order":0},{"forum_id":4111,"role":"chair","name":"白璐","affiliation":"北京师范大学人工智能学院","sort_order":0},{"forum_id":4111,"role":"guest","name":"魏哲巍","affiliation":"中国人民大学","sort_order":1},{"forum_id":4111,"role":"guest","name":"刘淇","affiliation":"中国科学技术大学","sort_order":2},{"forum_id":4111,"role":"guest","name":"江波","affiliation":"安徽大学","sort_order":3},{"forum_id":4111,"role":"guest","name":"钱宇华","affiliation":"山西大学","sort_order":4},{"forum_id":4111,"role":"guest","name":"黄华","affiliation":"北京师范大学人工智能学院","sort_order":5},{"forum_id":4111,"role":"guest","name":"陈恩红","affiliation":"中国科学技术大学","sort_order":6},{"forum_id":4111,"role":"guest","name":"于剑","affiliation":"北京交通大学","sort_order":7},{"forum_id":4111,"role":"guest","name":"石川","affiliation":"北京邮电大学","sort_order":8},{"forum_id":4111,"role":"guest","name":"曹飞龙","affiliation":"中国计量大学","sort_order":9},{"forum_id":4111,"role":"guest","name":"崔丽欣","affiliation":"中央财经大学","sort_order":10},{"forum_id":4111,"role":"guest","name":"李明","affiliation":"浙江师范大学","sort_order":11}],"host":"","image_url":"/api/forums/4111/cover"},{"id":4113,"title":"Panoramic Memory-Driven Personalized Agent Technology Forum","topic":"Unclassified","kind":"专题论坛","chair":"王茜莺、马思伟","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城2号馆-207","description":"Based on the important consensus of the February 2026 Xiuhu Conference, academia and industry have further recognized that \"memory\" is becoming the key infrastructure for personalized agents to move toward long-term intelligence. With the theme of \"Panoramic Memory Technology for Personalized Agents,\" this forum gathers national-level high-level talents, Distinguished Young Scholars, Changjiang Scholars, and founders of memory intelligence unicorn companies to jointly discuss the development trends and core challenges of the next-generation agent memory system.\n\nThe forum focuses on the evolution from \"local memory agents\" to \"panoramic memory agents.\" The forum focuses on the evolution from \"local memory agents\" to \"panoramic memory agents,\" exploring a unified memory system covering multimodality, multiple scenarios, and cross-time, achieving long-term understanding of user state, preferences, and goal evolution, and supporting agents' long-term cognition and proactive services. The forum will conduct discussions around key topics such as memory generation, edge-side evolution, cross-modal fusion, and memory-reasoning collaboration.","is_demo":0,"cover_url":"/api/forums/4113/cover?size=thumb","cover_alt":"Panoramic Memory-Driven Personalized Agent Technology Forum","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4113","source_checked_at":"2026-10-30","agenda_json":"[[\"Panoramic Memory: Top Ten Trends and Challenges in Agent Memory Technology in 2026\",\"王昊奋\"],[\"Multimodal Memory Systems for Large Model Agents\",\"张宁豫\"],[\"Making AI Remember You: On-Device Visual Memory and the Next Step for Personal AI\",\"Shawn Shen（沈俊潇）\"],[\"Self-Evolving Skill Accumulation and Efficiency Optimization for Personalized Agents\",\"韩军功\"],[\"Metis: From Stateless Large Models to Memory-Native Foundation Models\",\"熊飞宇\"],[\"AI Twin: Panoramic Memory Driving Agents Toward Intelligent Avatars\",\"李辉\"]]","code":"CNCC2026-4113","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城2号馆","room":"西博城2号馆-207","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4113","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"全景记忆驱动的个性化智能体技术论坛","original_topic":"未分类","original_description":"基于2026年2月秀湖会议的重要共识，学界与产业界进一步认识到，“记忆”正成为个性化智能体迈向长期智能的关键基础设施。本论坛以“个性化智能体的全景记忆技术”为主题，汇聚国家级高层次人才、杰青、长江学者及记忆智能独角兽企业创始人，共同探讨下一代智能体记忆体系的发展趋势与核心挑战 \n\n论坛聚焦从“局部记忆智能体”向“全景记忆智能体”的演进。论坛聚焦从“局部记忆智能体”向“全景记忆智能体”的演进，探索覆盖多模态、多场景、跨时间的统一记忆体系，实现对用户状态、偏好与目标演化的长期理解，支撑智能体长期认知与主动服务。论坛将围绕记忆生成、端侧演化、跨模态融合、记忆与推理协同等关键议题展开研讨。","people":[{"forum_id":4113,"role":"guest","name":"王昊奋","affiliation":"同济大学","sort_order":0},{"forum_id":4113,"role":"chair","name":"王茜莺","affiliation":"联想集团","sort_order":0},{"forum_id":4113,"role":"guest","name":"张宁豫","affiliation":"浙江大学","sort_order":1},{"forum_id":4113,"role":"co_chair","name":"马思伟","affiliation":"北京大学","sort_order":1},{"forum_id":4113,"role":"guest","name":"Shawn Shen（沈俊潇）","affiliation":"Memories.ai","sort_order":2},{"forum_id":4113,"role":"guest","name":"韩军功","affiliation":"清华大学自动化系","sort_order":3},{"forum_id":4113,"role":"guest","name":"熊飞宇","affiliation":"记忆张量（上海）科技有限公司","sort_order":4},{"forum_id":4113,"role":"guest","name":"李辉","affiliation":"联想集团","sort_order":5}],"host":"","image_url":"/api/forums/4113/cover"},{"id":4114,"title":"From Interaction to Symbiosis: Brain-Computer Interfaces and Embodied Intelligence","topic":"Robotics and Embodied Intelligence","kind":"专题论坛","chair":"梁臻","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-503","description":"In the context of \"data-driven aggregation, intelligent computing for the future,\" artificial intelligence is moving from data-driven approaches toward intelligent systems that integrate perception, cognition, and action. Embodied intelligence drives agents into the real physical world, while brain-computer interfaces provide a new information channel for machines to understand human intentions, cognitive states, and behavioral decisions, promoting the relationship between humans and machines to move from \"interaction\" toward \"symbiosis.\" This forum will focus on the frontier trends in the integrated development of brain-computer interfaces and embodied intelligence, and around key topics such as brain-inspired intelligence, brain-computer collaboration, embodied perception and decision-making, and human-machine integrated intelligence, discuss how brain science can empower next-generation intelligent systems, as well as the new opportunities and challenges brought by technologies such as large models, agents, and multimodal intelligence. At the same time, the forum will also exchange views on key issues such as generalization learning, human-machine collaboration, and safety and trustworthiness, promoting interdisciplinary integration in fields such as artificial intelligence, brain science, and robotics, and exploring new paradigms of intelligent computing in the era of human-machine symbiosis.","is_demo":0,"cover_url":"/api/forums/4114/cover?size=thumb","cover_alt":"From Interaction to Symbiosis: Brain-Computer Interfaces and Embodied Intelligence","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4114","source_checked_at":"2026-10-30","agenda_json":"[[\"A Preliminary Exploration of Embodied Intelligent Control Large Models for Industrial Vertical Domains\",\"阳春华\"],[\"From Learning Machines to Embodied Intelligence\",\"孙长银\"],[\"World-Model-Driven Cloud-Device Integrated Embodied Intelligence\",\"田永鸿\"],[\"First-Person Action Imitation and Controllable Generation\",\"李宏亮\"],[\"GSnet-Based Multi-Agent Collaborative Design Optimization System for Major Equipment\",\"梁骁俊\"]]","code":"CNCC2026-4114","forum_type":"专题论坛","format":null,"theme_code":"机器人与具身智能","theme_name":"机器人与具身智能","venue":"西博城5号馆","room":"西博城5号馆-503","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4114","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"从交互到共生：脑机接口与具身智能","original_topic":"机器人与具身智能","original_description":"在“数聚驱动，智算未来”的背景下，人工智能正由数据驱动迈向感知、认知与行动一体化的智能系统。具身智能推动智能体走向真实物理世界，脑机接口则为机器理解人类意图、认知状态和行为决策提供了全新的信息通道，推动人与机器的关系从“交互”迈向“共生”。本论坛将聚焦脑机接口与具身智能融合发展的前沿趋势，围绕脑启发智能、脑机协同、具身感知与决策、人机共融智能等关键议题，探讨脑科学如何赋能下一代智能系统，以及大模型、智能体、多模态智能等技术带来的新机遇与新挑战。同时，论坛还将围绕泛化学习、人机协同、安全可信等关键问题展开交流，促进人工智能、脑科学、机器人等领域的跨学科融合，探索人机共生时代的智能计算新范式。","people":[{"forum_id":4114,"role":"chair","name":"梁臻","affiliation":"深圳大学","sort_order":0},{"forum_id":4114,"role":"guest","name":"阳春华","affiliation":"中南大学","sort_order":0},{"forum_id":4114,"role":"guest","name":"孙长银","affiliation":"安徽大学","sort_order":1},{"forum_id":4114,"role":"guest","name":"田永鸿","affiliation":"北京大学","sort_order":2},{"forum_id":4114,"role":"guest","name":"李宏亮","affiliation":"电子科技大学","sort_order":3},{"forum_id":4114,"role":"guest","name":"梁骁俊","affiliation":"鹏城实验室","sort_order":4}],"host":"","image_url":"/api/forums/4114/cover"},{"id":4115,"title":"Information Confidentiality Confrontation in the AI Era","topic":"Unclassified","kind":"专题论坛","chair":"卿昱","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-115A九寨厅A","description":"This forum focuses on the digital and intelligent development of the confidentiality technology field, conducting technical discussions around addressing the confidentiality risk challenges brought by intelligence and AI-enabled confidentiality business. It aims to bring together experts from government, industry, academia, research, and application to share cutting-edge technological achievements of AI in the confidentiality field, and to provide ideas for building intelligent confidentiality technology development.","is_demo":0,"cover_url":"/api/forums/4115/cover?size=thumb","cover_alt":"Information Confidentiality Confrontation in the AI Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4115","source_checked_at":"2026-10-30","agenda_json":"[[\"Risks and Challenges That Artificial Intelligence Brings to Information Confidentiality\",\"吴世忠\"],[\"Several Dimensions of Large Model Agent Information Leakage\",\"虎嵩林\"],[\"Key Technologies of the ChatBIT Large Model and Intelligent Identification of Open-Source Intelligence Risks\",\"张华平\"],[\"AI Large Models Reshaping Cyber Offense and Defense and a New Line of Defense for Information Confidentiality\",\"陈庶樵\"],[\"Application of the HarmonyOS AI Security Foundation in the Field of Information Confidentiality\",\"付天福\"]]","code":"CNCC2026-4115","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-115A九寨厅A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4115","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"人工智能时代的信息保密对抗","original_topic":"未分类","original_description":"本论坛聚焦保密科技领域数智化发展，围绕应对智能化带来的保密风险挑战及人工智能赋能保密业务开展技术研讨，旨在汇聚政产学研用等专家分享人工智能在保密领域的前沿技术成果，为构建智能化的保密科技发展提供思路。","people":[{"forum_id":4115,"role":"chair","name":"卿昱","affiliation":"中关村实验室","sort_order":0},{"forum_id":4115,"role":"guest","name":"吴世忠","affiliation":"中关村实验室","sort_order":0},{"forum_id":4115,"role":"guest","name":"虎嵩林","affiliation":"中国科学院信息工程研究所","sort_order":1},{"forum_id":4115,"role":"guest","name":"张华平","affiliation":"北京理工大学","sort_order":2},{"forum_id":4115,"role":"guest","name":"陈庶樵","affiliation":"中关村实验室","sort_order":3},{"forum_id":4115,"role":"guest","name":"付天福","affiliation":"华为技术有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4115/cover"},{"id":4116,"title":"The 3rd On-Device Large Model Forum: Defining On-Device Artificial General Intelligence","topic":"Unclassified","kind":"专题论坛","chair":"刘知远","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-502","description":"The release of ChatGPT at the end of 2022 triggered global attention to large models. In addition to continuously scaling up models along the Scaling Law, as AI becomes deeply involved in its own R&D processes such as model architecture, training frameworks, and data synthesis, AI making AI (AI4AI) is driving large models to exhibit the Densing Law, in which capability density continues to increase. The superposition of the Densing Law and Moore's Law allows powerful intelligence to move from the cloud to the device: on-device intelligence is moving from \"being able to converse\" toward on-device agents that \"can perceive, reason, and act,\" initially possessing cloud-device collaboration and self-evolution capabilities, accelerating deployment in scenarios such as AI phones, AIPC, smart cockpits, smart wearables, smart homes, and even embodied robots, and increasingly becoming an indispensable intelligent foundation. This forum focuses on the frontier technologies and deployment practices of on-device large models, jointly envisioning the prospects and challenges of on-device intelligence.","is_demo":0,"cover_url":"/api/forums/4116/cover?size=thumb","cover_alt":"The 3rd On-Device Large Model Forum: Defining On-Device Artificial General Intelligence","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4116","source_checked_at":"2026-10-30","agenda_json":"[[\"Real-Time Omni-Modal Large Models Toward Human-Like Interaction\",\"姚远\"],[\"Efficient Construction of Large Models with High Capability Density\",\"肖朝军\"],[\"AI Manufacturing AI: Forge Engineering-Driven Chip-Level Training and Inference Customization for On-Device Large Models\",\"李宇轩\"],[\"VoxCPM: A Continuous Autoregressive Speech Generation Model\",\"曾国洋\"],[\"Large-Model Autonomous Agents Toward General Intelligence\",\"林衍凯\"]]","code":"CNCC2026-4116","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-502","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4116","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第三届端侧大模型论坛：定义端侧通用人工智能","original_topic":"未分类","original_description":"2022年底ChatGPT发布引发全球对大模型的关注。除沿规模定律（Scaling Law）将模型不断变大，随着AI深度参与模型架构、训练框架、数据合成等自身研发环节，AI制造AI（AI4AI）正推动大模型呈现能力密度持续增强的密度定律（Densing Law）。密度定律与摩尔定律相互叠加，让强大的智能从云端走向端侧：端侧智能正从\"能对话\"迈向\"能感知、能推理、能行动\"的端侧智能体，初步具备端云协同与自我进化能力，在AI手机、AIPC、智能座舱、智能穿戴、智能家居乃至具身机器人等场景中加速落地、日益成为不可或缺的智能底座。本论坛聚焦端侧大模型的前沿技术与落地实践，共同展望端侧智能的愿景与挑战。","people":[{"forum_id":4116,"role":"chair","name":"刘知远","affiliation":"清华大学","sort_order":0},{"forum_id":4116,"role":"guest","name":"姚远","affiliation":"清华大学","sort_order":0},{"forum_id":4116,"role":"guest","name":"肖朝军","affiliation":"清华大学","sort_order":1},{"forum_id":4116,"role":"guest","name":"李宇轩","affiliation":"面壁智能","sort_order":2},{"forum_id":4116,"role":"guest","name":"曾国洋","affiliation":"面壁智能","sort_order":3},{"forum_id":4116,"role":"guest","name":"林衍凯","affiliation":"中国人民大学","sort_order":4}],"host":"","image_url":"/api/forums/4116/cover"},{"id":4117,"title":"Opportunities and Challenges of Open-Source AI Infra in the Token Economy","topic":"Unclassified","kind":"专题论坛","chair":"章文嵩、章津楠","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-404","description":"As large models move from capability exploration to large-scale application, AI Infra is entering a new stage of development.\nOn the one hand, model scale, multimodal tasks, and agent applications continue to push up demands for computation, memory access, communication, and scheduling; on the other hand, the industry's evaluation criteria for AI infrastructure are gradually shifting from single-point computing power and model performance toward cost per Token, resource utilization efficiency, and the comprehensive cost of task completion.\nAgainst this background, how should AI Infra evolve? What new forms of collaboration will emerge among hardware architecture, cloud-native infrastructure, distributed inference, heterogeneous computing, and open-source software stacks?\nThis CNCC sub-forum, \"Opportunities and Challenges of AI Infra in the Token Economy,\" will invite multiple experts from industry, academia, and the open-source community to discuss from different perspectives, including cloud-native InfraServices, efficient deployment of large models, distributed inference, AI for Science, and Agentic AI Infra.","is_demo":0,"cover_url":"/api/forums/4117/cover?size=thumb","cover_alt":"Opportunities and Challenges of Open-Source AI Infra in the Token Economy","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4117","source_checked_at":"2026-10-30","agenda_json":"[[\"From Token Efficiency to Agent Productivity: Open-Source AI Infra Innovation for the Agent Era\",\"马涛\"],[\"llm-d: Building a Distributed Inference Plane Across Workloads, Modalities, Hardware, and Platforms\",\"张家驹\"],[\"Inference Practice for Heterogeneous Compute: Multi-Chip Adaptation and Efficient Execution in the FlagOS Ecosystem\",\"刘艳青\"],[\"Research on Efficient Deployment and Optimization Technologies for Large Models\",\"宋卓然\"],[\"From AI4S Models to Scientific Intelligence\",\"赵学庆\"],[\"Opportunities for Cloud-Native InfraServices in the Large Model Era\",\"章文嵩\"]]","code":"CNCC2026-4117","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-404","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4117","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"Token经济下的开源AI Infra机遇和挑战","original_topic":"未分类","original_description":"随着大模型从能力探索走向规模化应用，AI Infra 正在进入新的发展阶段。\n一方面，模型规模、多模态任务与智能体应用持续推高计算、访存、通信和调度需求；另一方面，行业对 AI 基础设施的评价标准，也正在从单点算力与模型性能，逐渐走向单位 Token 成本、资源利用效率以及任务完成的综合成本。\n在这样的背景下，AI Infra 应该如何演进？硬件架构、云原生基础设施、分布式推理、异构计算与开源软件栈之间又将形成怎样的新协作方式？\n本次 CNCC「Token 经济下的 AI Infra 的机遇与挑战」分论坛将邀请来自产业界、学术界及开源社区的多位专家，从云原生 InfraServices、大模型高效部署、分布式推理、科学智能到 Agentic AI Infra 等不同视角展开讨论。","people":[{"forum_id":4117,"role":"chair","name":"章文嵩","affiliation":"安托盟丘","sort_order":0},{"forum_id":4117,"role":"guest","name":"马涛","affiliation":"阿里巴巴","sort_order":0},{"forum_id":4117,"role":"co_chair","name":"章津楠","affiliation":"CCF开源发展技术委员会","sort_order":1},{"forum_id":4117,"role":"guest","name":"张家驹","affiliation":"红帽软件（北京）有限公司","sort_order":1},{"forum_id":4117,"role":"guest","name":"刘艳青","affiliation":"先进编译实验室","sort_order":2},{"forum_id":4117,"role":"guest","name":"宋卓然","affiliation":"上海交通大学","sort_order":3},{"forum_id":4117,"role":"guest","name":"赵学庆","affiliation":"沐曦集成电路（上海）股份有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4117/cover"},{"id":4120,"title":"The 3rd National Data Infrastructure Forum — AI-Integrated Trusted Data Spaces","topic":"Data Science and Data Spaces","kind":"专题论坛","chair":"高志鹏、杨阳","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-515","description":"In 2022, the Central Committee of the Communist Party of China and the State Council issued the \"Opinions on Accelerating the Building of a Unified National Market,\" explicitly proposing to \"accelerate the cultivation of the data element market.\" In 2026, the \"15th Five-Year Plan\" outline takes \"building an open, shared, and secure national integrated data market\" as one of the key tasks of Digital China construction. Under the guidance of the task of building a national integrated data market, how should the construction of China's data infrastructure balance efficiency and security? How can artificial intelligence be organically integrated with trusted data spaces? How can cross-border data flows be vigorously promoted under the premise of compliance? How can scientific and technological breakthroughs and industrial progress be achieved in tandem under policy guidance? Targeting these questions, this forum invites experts from the policy, academic, and industrial communities to discuss \"integrated and trusted data spaces,\" providing comprehensive and systematic reports and in-depth speculative discussion on the overall layout of national data infrastructure, compliant cross-border data flows, scientific and technological breakthroughs supporting industrial development, and key data technology achievements.","is_demo":0,"cover_url":"/api/forums/4120/cover?size=thumb","cover_alt":"The 3rd National Data Infrastructure Forum — AI-Integrated Trusted Data Spaces","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4120","source_checked_at":"2026-10-30","agenda_json":"[[\"Forum Background and Introduction of Speakers\",\"高志鹏\"],[\"Major Infrastructure for Cross-Border Data: Mechanisms and Practice\",\"陆志鹏\"],[\"Opportunities and Challenges for Data Spaces in the Intelligent Era\",\"黄罡\"],[\"Industrial Data Infrastructure: An Industrial Big Data Space Operating System for Discrete Industries\",\"刘驰\"],[\"Security and Privacy Challenges of Large Models from the Perspective of Trusted Data Spaces\",\"王志波\"],[\"Intelligent Data Governance for Cross-Domain Circulation: Opportunities and Challenges\",\"范举\"]]","code":"CNCC2026-4120","forum_type":"专题论坛","format":null,"theme_code":"数据科学与数据空间","theme_name":"数据科学与数据空间","venue":"西博城5号馆","room":"西博城5号馆-515","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4120","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第三届国家数据基础设施论坛——AI融合的一体化可信数据空间","original_topic":"数据科学与数据空间","original_description":"2022年，中共中央、国务院发布《关于加快建设全国统一大市场的意见》明确提出“加快培育数据要素市场”。2026年，“十五五”规划纲要将“建设开放共享安全的全国一体化数据市场”作为数字中国建设的重点任务之一。在全国一体化数据市场建设任务的牵引下，我国数据基础设施的建设该如何协调效率与安全的平衡？人工智能如何与可信数据空间有机结合？如何在合规的前提下大力促进数据跨境流通？如何在政策的牵引下实现科技攻关与产业进步协同并举？本论坛针对这些问题，邀请了政策界、学术界、产业界的专家共话“融合可信的一体化数据空间”，从国家数据基础设施整体布局、合规的跨境数据流动、科技攻关支撑产业发展、关键数据技术成果等方面进行全面系统的汇报，以及深刻的思辨讨论。","people":[{"forum_id":4120,"role":"chair","name":"高志鹏","affiliation":"北京邮电大学","sort_order":0},{"forum_id":4120,"role":"guest","name":"陆志鹏","affiliation":"横琴数据运河科技有限公司","sort_order":0},{"forum_id":4120,"role":"co_chair","name":"杨阳","affiliation":"电子科技大学","sort_order":1},{"forum_id":4120,"role":"guest","name":"黄罡","affiliation":"北京大学","sort_order":1},{"forum_id":4120,"role":"guest","name":"刘驰","affiliation":"北京理工大学","sort_order":2},{"forum_id":4120,"role":"guest","name":"王志波","affiliation":"浙江大学","sort_order":3},{"forum_id":4120,"role":"guest","name":"范举","affiliation":"中国人民大学","sort_order":4}],"host":"","image_url":"/api/forums/4120/cover"},{"id":4128,"title":"Spatial Intelligence Forum: Frontier Exploration of 3D Real-Scene Reconstruction and AIGC Generation","topic":"Vision, Multimodality and World Models","kind":"专题论坛","chair":"周昆","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-508","description":"3D real-scene reconstruction and generation technology has become a highly prominent core direction in the field of spatial intelligence. In recent years, technologies such as 3D Gaussian Splatting (3DGS) and 3D AIGC have iterated rapidly, achieving significant breakthroughs in accuracy, efficiency, and realism, and have been deployed in key fields such as digital twins, smart cities, and digital humans. At the same time, related theoretical innovations and engineering practices continue to emerge, with research directions covering high-fidelity 3D generation, efficient reconstruction and rendering of large scenes, and multimodal 3D generation foundation models. Against this background, this forum conducts exchanges and discussions from dimensions such as technical principles, algorithm innovation, and system implementation, showcases frontier progress, focuses on key challenges, deepens industry-academia-research collaboration, and continuously promotes the high-quality development of 3D real-scene reconstruction and generation technology in the era of spatial intelligence.","is_demo":0,"cover_url":"/api/forums/4128/cover?size=thumb","cover_alt":"Spatial Intelligence Forum: Frontier Exploration of 3D Real-Scene Reconstruction and AIGC Generation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4128","source_checked_at":"2026-10-30","agenda_json":"[[\"Visual Localization with Lightweight Structured Maps\",\"申抒含\"],[\"Digital Space Modeling Driven by Dual Engines: Explicit Geometric Representation and Implicit Generative Models\",\"高林\"],[\"Research on Dual-Scale 3D Photorealistic Representation with Gaussian+X\",\"邵天甲\"],[\"3D Foundation Models: From 3D Reconstruction to General Image Geometric Representation\",\"姚遥\"],[\"Efficient 3D Content Generation and Controllable Editing\",\"穆太江\"]]","code":"CNCC2026-4128","forum_type":"专题论坛","format":null,"theme_code":"视觉、多模态与世界模型","theme_name":"视觉、多模态与世界模型","venue":"西博城5号馆","room":"西博城5号馆-508","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4128","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"空间智能技术论坛：三维实景重建与 AIGC 生成前沿探索","original_topic":"视觉、多模态与世界模型","original_description":"三维实景重建与生成技术已成为空间智能领域备受瞩目的核心方向。近年来，三维高斯泼溅（3DGS）、三维AIGC等技术快速迭代，在精度、效率与真实感上实现显著突破，已在数字孪生、智慧城市、数字人等关键领域落地应用。与此同时，相关理论创新与工程实践持续涌现，研究方向涵盖高保真三维生成、大场景高效重建与绘制、多模态三维生成大模型等。在此背景下，本论坛从技术原理、算法创新、系统实现等维度展开交流研讨，展示前沿进展，聚焦关键挑战，深化产学研协同，持续推动空间智能时代三维实景重建与生成技术高质量发展。","people":[{"forum_id":4128,"role":"chair","name":"周昆","affiliation":"浙江大学","sort_order":0},{"forum_id":4128,"role":"guest","name":"申抒含","affiliation":"中国科学院自动化研究所","sort_order":0},{"forum_id":4128,"role":"guest","name":"高林","affiliation":"中国科学院计算技术研究所","sort_order":1},{"forum_id":4128,"role":"guest","name":"邵天甲","affiliation":"浙江大学","sort_order":2},{"forum_id":4128,"role":"guest","name":"姚遥","affiliation":"南京大学","sort_order":3},{"forum_id":4128,"role":"guest","name":"穆太江","affiliation":"清华大学","sort_order":4}],"host":"","image_url":"/api/forums/4128/cover"},{"id":4139,"title":"Breaking Compute Anxiety: How Compute-in-Memory Integration Can Ignite the Efficiency Revolution in the Agent Era","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"张伟丰、何哲陟","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-406","description":"As AI evolves from a passive assistant into autonomous agents, the system bottleneck has shifted from computing power to data movement. Compute-in-memory integration performs computation directly at the data residence location, fundamentally eliminating the energy consumption and latency bottlenecks caused by data movement in the traditional von Neumann architecture, and providing low-latency, high-energy-efficiency support for agents' multi-step reasoning, dynamic knowledge graph maintenance, and tool invocation. At the same time, agents also need to maintain stable throughput under dynamic loads, including real-time retrieval, memory updates, and streaming output decisions. Compute-in-memory integration architectures naturally fit these needs, making edge-resident agents and cloud-edge collaborative hybrid intelligent computing technically and economically feasible. Compute-in-memory integration improves efficiency and reduces costs, and can not only respond to the current challenge of high storage costs, but also has long-term strategic value for intelligent computing architecture innovation. This forum gathers academic and industry experts to focus on frontier technologies of compute-in-memory integration, AI-native architectures, efficiency innovation, and ecosystem building, supporting the national grand strategy of the intelligent computing revolution.","is_demo":0,"cover_url":"/api/forums/4139/cover?size=thumb","cover_alt":"Breaking Compute Anxiety: How Compute-in-Memory Integration Can Ignite the Efficiency Revolution in the Agent Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4139","source_checked_at":"2026-10-30","agenda_json":"[[\"Compute-Storage Fusion Pooled System for Large Model Inference\",\"肖利民\"],[\"AI and Compute-in-Memory Integration\",\"戴瑾\"],[\"Hierarchical \\\"Network-Storage-Compute\\\" Fusion Technology for Large Model Scenarios\",\"孙广宇\"],[\"High-Density CMOS In-Memory Computing Circuits and Architectures\",\"李学清\"],[\"DNA Molecular Computing and Storage: Depicting Future Green Computing Power with Biotechnology\",\"王飞\"]]","code":"CNCC2026-4139","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-406","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4139","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"打破算力焦虑：存算融合如何引爆智能体时代的效能革命","original_topic":"新型计算与网络架构","original_description":"随着AI从被动助手演进为自主智能体，系统瓶颈已从算力转向数据搬运。存算融合在数据驻留位置直接执行计算，从根本上消除统冯·诺依曼架构中数据搬运所带来的能耗和延迟瓶颈，为智能体多步推理、动态知识图谱维护与工具调用提供低延迟、高能效支撑。同时，智能体还需在动态负载下保持稳定吞吐，包括实时检索、更新记忆与流式输出决策。而存算融合架构天然适配这些需求，使边缘常驻智能体与云边协同混合智算在技术与经济上成为现实。存算融合提效降本，不仅能回应存储成本高企的当下挑战，更具智算架构创新的长期战略价值。本论坛汇聚学术与产业专家，聚焦存算融合前沿技术、AI原生架构、效能创新及生态建设，助力国家智算革命大战略。","people":[{"forum_id":4139,"role":"chair","name":"张伟丰","affiliation":"联想（北京）有限公司","sort_order":0},{"forum_id":4139,"role":"guest","name":"肖利民","affiliation":"北京航空航天大学","sort_order":0},{"forum_id":4139,"role":"co_chair","name":"何哲陟","affiliation":"上海交通大学","sort_order":1},{"forum_id":4139,"role":"guest","name":"戴瑾","affiliation":"北京超弦存储器研究院","sort_order":1},{"forum_id":4139,"role":"guest","name":"孙广宇","affiliation":"北京大学","sort_order":2},{"forum_id":4139,"role":"guest","name":"李学清","affiliation":"清华大学","sort_order":3},{"forum_id":4139,"role":"guest","name":"王飞","affiliation":"上海交通大学","sort_order":4},{"forum_id":4139,"role":"guest","name":"赵振伟","affiliation":"长鑫科技","sort_order":5},{"forum_id":4139,"role":"guest","name":"蒋超","affiliation":"联想集团","sort_order":6}],"host":"","image_url":"/api/forums/4139/cover"},{"id":4144,"title":"3rd AI4S Forum — AI for Science and Social Intelligence: Dual Drivers, Mutual Inspiration","topic":"AI4Science","kind":"专题论坛","chair":"王新","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-506","description":"This forum is the third AI for Science forum that Fudan University has held consecutively at CNCC. Under the theme \"AI for Science and Social Intelligence: Dual Drivers, Mutual Inspiration,\" it aims to explore in depth the underlying logic of bidirectional empowerment and collaborative evolution between AI for Science and AI for Social Science. At present, AI for Science is reconstructing natural-science research paradigms such as life sciences and drug discovery at an unprecedented speed, while social intelligence demonstrates strong explanatory and decision-making power in policy simulation, public governance, and the analysis of complex social systems. However, the two have long belonged to different academic fields, and their insights have not been fully exchanged. This forum is committed to breaking down this barrier: it focuses on how methodologies of AI for Science, such as scientific foundation models and agent-based simulation, can provide new tools and new perspectives for modeling social behavior, and it also emphasizes how ethical frameworks, institutional design, and systems thinking in social intelligence can feed back into AI for Science, ensuring that its development is beneficial, controllable, and sustainable. The forum invites top experts in biology, medicine, policy, management, and other fields to give keynote reports, and to engage in in-depth reflection and open exchange around key questions of mutual inspiration between the two intelligence systems, striving to promote the transition of natural sciences and social sciences from parallel development to collaborative progress, and to contribute interdisciplinary wisdom to human well-being in the intelligent era.","is_demo":0,"cover_url":"/api/forums/4144/cover?size=thumb","cover_alt":"3rd AI4S Forum — AI for Science and Social Intelligence: Dual Drivers, Mutual Inspiration","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4144","source_checked_at":"2026-10-30","agenda_json":"[[\"Main Ways AI Intervenes in Social Science Research and the Construction of Intelligent Innovation Engines\",\"孟庆国\"],[\"BrainFuse: A Large-Scale Fine-Grained Neural Twin Platform\",\"李国齐\"],[\"\\\"Social Model\\\": Toward a Technical Platform for Simulating and Deducting Core Real-World Social Problems\",\"唐世平\"],[\"Virtual Cell: AI-Driven Life Simulation\",\"彭玺\"],[\"A New Ecosystem for Individual (OPC) Innovation, Entrepreneurship, and Creativity Based on a Social \\\"World Model\\\"\",\"吕鹏\"]]","code":"CNCC2026-4144","forum_type":"专题论坛","format":null,"theme_code":"AI4Science","theme_name":"AI4Science","venue":"西博城5号馆","room":"西博城5号馆-506","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4144","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第三届科学智能（AI4S）论坛——科学智能与社会智能：双轮驱动·互启共进","original_topic":"AI4Science","original_description":"本论坛是复旦大学在CNCC上连续举办的第三届科学智能论坛，以“科学智能与社会智能：双轮驱动·互启共进”为主题，旨在深入探讨AI for Science与AI for Social Science之间双向赋能与协同演进的深层逻辑。当前，科学智能正以前所未有的速度重构生命科学、药物研发等自然科研范式，而社会智能则在政策模拟、公共治理和社会复杂系统分析中展现强大解释力与决策力，然而两者长期分属不同学术场域，彼此洞见未能充分交汇。本论坛致力于打破这一壁垒，既关注科学大模型、智能体仿真等科学智能的方法论，如何为社会行为建模提供新工具与新视角，也强调社会智能中的伦理框架、制度设计和系统思维如何反哺科学智能，确保其发展向善、可控且可持续。论坛邀请生物、医药、政策、管理等领域的顶尖专家做主旨报告，围绕两大智能体系相互启发的关键问题展开深度思辨与开放交流，力求推动自然科学与社会科学从平行发展走向协同共进，共同为智能时代的人类福祉贡献跨学科智慧。","people":[{"forum_id":4144,"role":"chair","name":"王新","affiliation":"复旦大学","sort_order":0},{"forum_id":4144,"role":"guest","name":"颜波","affiliation":"复旦大学","sort_order":0},{"forum_id":4144,"role":"guest","name":"孟庆国","affiliation":"清华大学","sort_order":1},{"forum_id":4144,"role":"guest","name":"李国齐","affiliation":"中国科学院自动化研究所","sort_order":2},{"forum_id":4144,"role":"guest","name":"唐世平","affiliation":"复旦大学","sort_order":3},{"forum_id":4144,"role":"guest","name":"彭玺","affiliation":"四川大学","sort_order":4},{"forum_id":4144,"role":"guest","name":"吕鹏","affiliation":"北京师范大学、重庆人工智能学院","sort_order":5}],"host":"","image_url":"/api/forums/4144/cover"},{"id":4151,"title":"Embodied and Co-Embodied Intelligence with Human-Machine-Environment Collaboration","topic":"Robotics and Embodied Intelligence","kind":"专题论坛","chair":"王涌天","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-509","description":"With the rapid development of embodied intelligence technology, the research focus has shifted from capability improvement of a single body to deep collaboration among humans, machines, and the environment, giving rise to new paradigms such as \"co-embodied intelligence.\" This forum closely follows this frontier trend and explores the leapfrog development from individual intelligence to collaborative intelligence and from embodied operation to co-embodied symbiosis, with reports organized around the following four aspects: ① 3D environment perception and reconstruction for complex scenarios, using computer vision and generative technologies to solve the problem of precise cognition; ② building embodied data collection and simulation platforms to break through the bottleneck of embodied intelligence data acquisition; ③ highly generalizable and robust embodied foundation models and decision architectures, combining reinforcement learning to endow agents with autonomous planning and evolution capabilities; ④ exploring human-machine collaborative intelligence and environment interaction mechanisms to achieve efficient collaboration between human and machine brains and the construction of co-embodied intelligence.","is_demo":0,"cover_url":"/api/forums/4151/cover?size=thumb","cover_alt":"Embodied and Co-Embodied Intelligence with Human-Machine-Environment Collaboration","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4151","source_checked_at":"2026-10-30","agenda_json":"[[\"A Causal Perspective on Human-Computer Interaction in the Era of Large Models: From Interactive Learning to User World Models\",\"喻纯\"],[\"Research Progress on Causal Rehearsal World Models for Decision-Making\",\"俞扬\"],[\"Co-Embodied Intelligence: A New Paradigm for Collaborative Human-Machine Embodied Intelligence\",\"陆峰\"],[\"Embodied Intelligence Manipulation Learning with Data-Perception-Decision Synergy\",\"施柏鑫\"],[\"Interaction-Aware Human-Object-Scene Reconstruction and Generation\",\"张鸿文\"]]","code":"CNCC2026-4151","forum_type":"专题论坛","format":null,"theme_code":"机器人与具身智能","theme_name":"机器人与具身智能","venue":"西博城5号馆","room":"西博城5号馆-509","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4151","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"人-机-环境协同的具身与共身智能","original_topic":"机器人与具身智能","original_description":"随着具身智能技术的飞速发展，研究重心已从单一本体的能力提升转向人、机、环境的深度协同，并衍生出“共身智能”等新范式。本论坛紧扣该前沿趋势，探讨从个体智能向协作智能、从具身操作向共身共生的跨越式发展，围绕以下四个方面开展报告：①面向复杂场景的三维环境感知与重构，利用计算机视觉与生成式技术解决精准认知问题；②构建具身数据采集与仿真平台，突破具身智能数据获取瓶颈；③高泛化、高鲁棒性的具身大模型与决策架构，结合强化学习赋予智能体自主规划与进化能力；④探索人机协同的智能与环境交互机制，实现人机双脑的高效协作与共身智能构建。","people":[{"forum_id":4151,"role":"chair","name":"王涌天","affiliation":"北京理工大学","sort_order":0},{"forum_id":4151,"role":"guest","name":"喻纯","affiliation":"清华大学","sort_order":0},{"forum_id":4151,"role":"guest","name":"俞扬","affiliation":"南京大学","sort_order":1},{"forum_id":4151,"role":"guest","name":"陆峰","affiliation":"北京航空航天大学","sort_order":2},{"forum_id":4151,"role":"guest","name":"施柏鑫","affiliation":"北京大学","sort_order":3},{"forum_id":4151,"role":"guest","name":"张鸿文","affiliation":"北京师范大学","sort_order":4}],"host":"","image_url":"/api/forums/4151/cover"},{"id":4152,"title":"Youth Power · Intelligent Frontiers: AI Innovation and Xiaomi Technology Collaboration Ecosystem","topic":"Unclassified","kind":"专题论坛","chair":"王斌","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城2号馆-204","description":"As large models move from the laboratory to thousands of industries, AI is undergoing a critical leap from algorithmic breakthroughs to industrial implementation. As a leader in \"human-car-home full-ecosystem\" smart living, Xiaomi has always resonated with the academic frontier—from embodied intelligence to multimodal perception, from edge chips to cloud large models, the depth and breadth of technological collaboration continue to expand.\n\nThis forum focuses on the most dynamic young scholar force in the AI field, gathering young research elites from top domestic universities to share the latest research results and industrial implementation practices around frontier directions such as large models, embodied intelligence, and AI for sicience.","is_demo":0,"cover_url":"/api/forums/4152/cover?size=thumb","cover_alt":"Youth Power · Intelligent Frontiers: AI Innovation and Xiaomi Technology Collaboration Ecosystem","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4152","source_checked_at":"2026-10-30","agenda_json":"[[\"GUI Agents: From Environment Interaction to Autonomous Planning\",\"黄萱菁\"],[\"Xiaomi Technology Collaboration and Technology Ecosystem\",\"陈虹雨\"],[\"From \\\"Trial-and-Error Driven\\\" to \\\"Autonomous Evolution\\\": Multi-Agent and Large Models Reshaping Materials R&D Paradigms\",\"孟二利\"],[\"Research on Efficient Inference and Multimodal Methods for Agents\",\"李祖超\"],[\"Research on Edge Large Model Inference Acceleration Based on Dataflow Technology\",\"范志华\"]]","code":"CNCC2026-4152","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城2号馆","room":"西博城2号馆-204","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4152","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"青年力量·智能前沿-AI创新与小米技术合作生态","original_topic":"未分类","original_description":"当大模型从实验室走向千行百业，AI正经历从算法突破到产业落地的关键一跃。小米作为\"人车家全生态\"智能生活的引领者，始终与学术前沿同频共振——从具身智能到多模态感知，从端侧芯片到云端大模型，技术合作的深度与广度持续拓展。\n\n本论坛聚焦AI领域最具活力的青年学者力量，汇聚来自国内顶尖高校的青年科研精英，围绕大模型、具身智能、AI for sicience等前沿方向，分享最新研究成果与产业落地实践。","people":[{"forum_id":4152,"role":"chair","name":"王斌","affiliation":"小米集团","sort_order":0},{"forum_id":4152,"role":"guest","name":"黄萱菁","affiliation":"复旦大学","sort_order":0},{"forum_id":4152,"role":"guest","name":"陈虹雨","affiliation":"小米集团","sort_order":1},{"forum_id":4152,"role":"guest","name":"孟二利","affiliation":"小米集团","sort_order":2},{"forum_id":4152,"role":"guest","name":"李祖超","affiliation":"武汉大学","sort_order":3},{"forum_id":4152,"role":"guest","name":"范志华","affiliation":"中国科学院计算技术研究所","sort_order":4}],"host":"","image_url":"/api/forums/4152/cover"},{"id":4153,"title":"Extended Reality and World Models: Building an Interactive Digital World","topic":"Vision, Multimodality and World Models","kind":"专题论坛","chair":"杨旭波","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-507","description":"Extended reality is a concentrated embodiment of the integration of the 3D digital world, the physical world, and the human world, while world models inject into it a causal engine capable of reasoning and prediction. The two form a bidirectional, mutually reinforcing evolutionary loop: XR interaction data can provide scarce training fuel for world models, physical feedback becomes the ultimate test ground for verifying model predictions, and millisecond-level real-time requirements force models to move from offline to lightweight; at the same time, the causal reasoning and spatiotemporal prediction capabilities of world models continuously enhance the immersion, interactivity, and imagination of extended reality.\n\nThis forum brings together scholars in related fields such as 3D perception, extended reality, world models, physical simulation, and embodied intelligence, focusing mainly on the following topics: 1) Integrated perception — how can virtual information and physical perception inputs be integrated to form a complete understanding and reconstruction of 3D space? How can world models use spatiotemporal prediction and causal reasoning capabilities to complete and correct occluded or missing information at the perception level, enhancing the immersion of extended reality? How can human embodied cognition and machine perception complement each other and provide richer training signals for world models? What perception-level challenges are involved? 2) Virtual-real interaction — how can virtual content and the physical world establish consistent 3D physical interaction rules? How can world models anticipate human interaction behavior and improve the naturalness and efficiency of 3D interaction? How can 3D interaction data in XR enable world models to more accurately understand human intentions and adapt to human interaction behavior patterns? What interaction-level challenges are involved? 3) Co-evolution — how can virtual simulation and physical feedback form a continuously iterating optimization loop? How can scenario generation and physical simulation verification through world models allow users to conceive hypothetical scenarios and observe their evolution in real time, driving the evolution of extended reality experiences and the enhancement of imagination? And how can deviations and successful experiences in physical execution be fed back to world models, continuously strengthening their prediction and reasoning capabilities through testing in real scenarios? What imagination-level challenges are involved?\n\nThe forum aims to explore the deep integration and collaborative development of extended reality and world models, driving the experiential evolution of extended reality and the continuous enhancement of world models through a bidirectional loop.","is_demo":0,"cover_url":"/api/forums/4153/cover?size=thumb","cover_alt":"Extended Reality and World Models: Building an Interactive Digital World","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4153","source_checked_at":"2026-10-30","agenda_json":"[[\"Physical AI-Driven Digital World Construction: From Physical Simulation to Immersive Interaction\",\"杨旭波\"],[\"Mechanism-Data Fusion World Models and Industrial Applications\",\"徐凯\"],[\"3D Reconstruction and Generation and 3D World Models\",\"章国锋\"],[\"Construction of 4D Dynamic World Models\",\"戴玉超\"],[\"Mechanism-Data Fusion World Models and Industrial Applications\",\"赵航\"],[\"World Models: From Geometric Appearance to Physical Semantics\",\"杨禹雪\"]]","code":"CNCC2026-4153","forum_type":"专题论坛","format":null,"theme_code":"视觉、多模态与世界模型","theme_name":"视觉、多模态与世界模型","venue":"西博城5号馆","room":"西博城5号馆-507","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4153","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"扩展现实与世界模型：构建可交互的数字世界","original_topic":"视觉、多模态与世界模型","original_description":"扩展现实是3D数字世界、物理世界与人类世界三界融合的集中体现，世界模型则为其注入可推理、可预测的因果引擎。两者构成双向互促的进化闭环：XR交互数据可为世界模型提供稀缺训练燃料，物理反馈成为检验模型预测的终极考场，毫秒级实时需求倒逼模型从离线走向轻量化；同时，世界模型的因果推理与时空预测能力，持续提升扩展现实的沉浸感、交互性与构想力。\n\n本论坛汇聚三维感知、扩展现实、世界模型、物理仿真与具身智能等相关领域学者，主要聚焦以下议题：1）融合感知——虚拟信息与物理感知输入如何融合，形成对三维空间的完整理解与重建？世界模型如何利用时空预测与因果推理能力，在感知层面对遮挡、缺失信息进行补全与纠偏，增强扩展现实的沉浸感？人的具身认知与机器感知如何互补，并为世界模型提供更丰富的训练信号？其中面临哪些感知层面的挑战？2）虚实交互——虚拟内容与物理世界如何建立一致的3D物理交互规则？世界模型如何预判人的交互行为，提升三维交互的自然性与高效性？XR中的三维交互数据如何使世界模型更准确地理解人类意图、适应人类交互行为模式？其中面临哪些交互层面的挑战？3）协同进化——虚拟推演与物理反馈如何形成持续迭代的优化闭环？如何通过世界模型的情景生成与物理仿真验证，使用户得以构想假设情景并实时观察其演化，驱动扩展现实的体验进化与构想力提升？又如何将物理执行中的偏差与成功经验反馈给世界模型，通过真实场景的检验不断强化其预测与推理能力？其中面临哪些构想力层面的挑战？\n\n论坛旨在探讨扩展现实与世界模型的深度融合与协同发展，通过双向闭环驱动扩展现实的体验进化与世界模型的持续增强。","people":[{"forum_id":4153,"role":"chair","name":"杨旭波","affiliation":"上海交通大学","sort_order":0},{"forum_id":4153,"role":"guest","name":"徐凯","affiliation":"中国科学院工业人工智能研究所","sort_order":0},{"forum_id":4153,"role":"guest","name":"高阳","affiliation":"北京航空航天大学","sort_order":1},{"forum_id":4153,"role":"guest","name":"章国锋","affiliation":"浙江大学","sort_order":2},{"forum_id":4153,"role":"guest","name":"戴玉超","affiliation":"西北工业大学电子信息学院","sort_order":3},{"forum_id":4153,"role":"guest","name":"赵航","affiliation":"中国科学院工业人工智能研究所","sort_order":4},{"forum_id":4153,"role":"guest","name":"杨禹雪","affiliation":"中国科学院自动化研究所","sort_order":5},{"forum_id":4153,"role":"guest","name":"汪淼","affiliation":"北京航空航天大学","sort_order":6}],"host":"","image_url":"/api/forums/4153/cover"},{"id":4156,"title":"Ubiquitous Mobile Computing Empowering the Low-Altitude Economy—The 7th Mobile Computing in Ubiquitous Intelligent Environments Debate Forum","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"赵亮、徐乐西","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-108邛海","description":"Ubiquitous Mobile Computing Empowering the Low-Altitude Economy—The 7th Mobile Computing in Ubiquitous Intelligent Environments Debate Forum | CNCC\nThe low-altitude economy is moving from single-point drone applications to a new stage of air-ground collaboration, integrated communication-sensing-computing and swarm intelligence, imposing higher requirements on network access, edge computing, intelligent perception, covert communication and collaborative decision-making. Ubiquitous mobile computing emphasizes continuous connectivity, real-time perception and collaborative computing between agents and the environment, and is an important technological foundation supporting the large-scale operation of low-altitude scenarios. At present, low-altitude applications still face challenges such as limited air-ground links, explosive growth of multimodal data, edge computing power constraints, and human-machine collaborative decision-making in complex threat environments; industrial deployment also requires communication, computing, navigation, intelligent terminals and industry applications to form a collaborative ecosystem. This forum focuses on how mobile computing in ubiquitous intelligent environments can empower the low-altitude economy, and invites experts in communication networks, cyberspace security, wireless communication, edge intelligence, semantic communication and low-altitude industry applications to discuss from the perspectives of operator practices, low-altitude prevention and control, covert communication, air-ground collaborative intelligent computing and low-altitude network solutions, jointly deliberating the key architectures, technical paths and industry directions for a ubiquitous intelligent mobile computing system to support the development of the low-altitude economy.","is_demo":0,"cover_url":"/api/forums/4156/cover?size=thumb","cover_alt":"Ubiquitous Mobile Computing Empowering the Low-Altitude Economy—The 7th Mobile Computing in Ubiquitous Intelligent Environments Debate Forum","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4156","source_checked_at":"2026-10-30","agenda_json":"[[\"Communication + AI + Low-Altitude: An Operator's Practice and Reflections\",\"唐雄燕\"],[\"Human-Machine Collaborative Command and Decision-Making for Low-Altitude Defense: Challenges and Key Pathways\",\"王震\"],[\"Theory and Methods for Low-Altitude Intelligent Covert Communication\",\"赵楠\"],[\"Semantics-Driven Air-Ground Collaborative Intelligent Computing Methods for the Low-Altitude Economy\",\"邢焕来\"],[\"Low-Altitude Network Solutions and Applications\",\"赵永茂\"]]","code":"CNCC2026-4156","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"天府国际会议中心","room":"天府国际会议中心-108邛海","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4156","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"泛在移动计算赋能低空经济—第七届泛在智能环境下的移动计算思辨论坛","original_topic":"新型计算与网络架构","original_description":"泛在移动计算赋能低空经济——第七届泛在智能环境下的移动计算思辨论坛 | CNCC\n低空经济正在从单点无人机应用走向空地协同、通感算融合和群体智能的新阶段，对网络接入、边缘计算、智能感知、隐蔽通信与协同决策提出了更高要求。泛在移动计算强调智能体与环境之间的连续连接、实时感知和协同计算，是支撑低空场景规模化运行的重要技术底座。当前，低空应用仍面临空地链路受限、多模态数据爆发、边缘算力约束、复杂威胁环境下的人机协同决策等挑战，产业落地也需要通信、计算、导航、智能终端与行业应用形成协同生态。本论坛聚焦泛在智能环境下的移动计算如何赋能低空经济，邀请通信网络、网络空间安全、无线通信、边缘智能、语义通信与低空产业应用等方向专家，从运营商实践、低空防控、隐蔽通信、空地协同智能计算和低空网络解决方案等角度展开讨论，共同思辨泛在智能移动计算体系支撑低空经济发展的关键架构、技术路径与产业方向。","people":[{"forum_id":4156,"role":"chair","name":"赵亮","affiliation":"沈阳航空航天大学","sort_order":0},{"forum_id":4156,"role":"guest","name":"唐雄燕","affiliation":"中国联合网络通信有限公司研究院","sort_order":0},{"forum_id":4156,"role":"co_chair","name":"徐乐西","affiliation":"中国联合网络通信有限公司研究院","sort_order":1},{"forum_id":4156,"role":"guest","name":"王震","affiliation":"西北工业大学","sort_order":1},{"forum_id":4156,"role":"guest","name":"赵楠","affiliation":"大连理工大学","sort_order":2},{"forum_id":4156,"role":"guest","name":"邢焕来","affiliation":"西南交通大学","sort_order":3},{"forum_id":4156,"role":"guest","name":"赵永茂","affiliation":"沈阳美行科技股份有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4156/cover"},{"id":4160,"title":"Low-Precision Computing Forum","topic":"Unclassified","kind":"专题论坛","chair":"弋飞","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-107A三星堆A","description":"As the parameter scale, context length and computational demands of large models continue to grow, compute power, storage, bandwidth and energy consumption are becoming important factors constraining the further development of artificial intelligence. By adopting 8-bit, 4-bit or even lower bit-width data representations, low-precision computing can effectively reduce the storage and movement overhead of model weights, activations and intermediate data, and improve computational throughput and energy efficiency, and is becoming an important research direction for new-generation artificial intelligence computing.\nHowever, low precision is not merely about reducing a few bits; as bit widths continue to decrease, the contradictions among model accuracy, numerical dynamic range, quantization error, training stability and hardware execution efficiency become more prominent. How to design new low-precision data formats better suited to AI data distributions, how to achieve coordination among training, inference and quantization algorithms, and how to truly translate theoretical bit-width advantages into end-to-end performance gains through compilation, operators and architecture design—these remain questions with substantial room for in-depth study.\nThis forum will focus on the academic frontiers of low-precision computing, covering low-precision numerical representations and data formats, low-bit model training and inference, quantization algorithms, software-hardware co-design, efficient operators and architectures, low-precision computing systems and data transmission, and other directions. Experts and scholars in related fields in China will be invited to share the latest research results and jointly discuss the fundamental issues, key technologies and research trends worthy of attention in the future of low-precision computing.","is_demo":0,"cover_url":"/api/forums/4160/cover?size=thumb","cover_alt":"Low-Precision Computing Forum","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4160","source_checked_at":"2026-10-30","agenda_json":"[[\"Low-Precision Operator Development and Performance Optimization\",\"陆璐\"],[\"Numerical Precision of Large Models: Misconceptions, Principles, and Bottlenecks\",\"尚笠\"],[\"Quantization Compression and Efficient Deployment of Large Models\",\"王培松\"],[\"Hardware-Software Co-Optimization of Low-Precision Data Formats for AI Computing\",\"鲁金铭\"],[\"Extremely Low-Bit Large Model Inference\",\"冷静文\"],[\"Toward Extremely Low-Precision Foundation Models: From Numerics to Tensors to Temporal Sequences\",\"秦浩桐\"],[\"Introduction to the Hifloat Format and Training/Inference Practice\",\"李思坦\"]]","code":"CNCC2026-4160","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-107A三星堆A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":7,"forum_id":"CNCC2026-4160","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"低精度计算论坛","original_topic":"未分类","original_description":"随着大模型参数规模、上下文长度和计算需求持续增长，算力、存储、带宽与能耗正在成为制约人工智能进一步发展的重要因素。通过采用8bit、4bit乃至更低位宽的数据表示，低精度计算能够有效减少模型权重、激活值及中间数据的存储与搬运开销，并提升计算吞吐与能效，正在成为新一代人工智能计算的重要研究方向。\n但低精度不只是减少几个bit，位宽不断降低后，模型精度、数值动态范围、量化误差、训练稳定性以及硬件执行效率之间的矛盾更加突出。如何设计更适合AI数据分布的新型低精度数据格式，如何实现训练、推理和量化算法协同，如何通过编译、算子与体系结构设计将理论上的位宽优势真正转化为端到端性能收益，仍存在大量值得深入研究的问题。\n本论坛将面向低精度计算的学术前沿，围绕低精度数值表示与数据格式、低比特模型训练与推理、量化算法、软硬件协同设计、高效算子与体系结构、低精度计算系统与数据传输等方向展开交流，邀请国内相关领域专家学者分享最新研究成果，并共同探讨低精度计算未来值得关注的基础问题、关键技术与研究趋势。","people":[{"forum_id":4160,"role":"chair","name":"弋飞","affiliation":"华为技术有限公司","sort_order":0},{"forum_id":4160,"role":"guest","name":"陆璐","affiliation":"华南理工大学","sort_order":0},{"forum_id":4160,"role":"guest","name":"黄朝春","affiliation":"华为","sort_order":1},{"forum_id":4160,"role":"guest","name":"尚笠","affiliation":"复旦大学","sort_order":2},{"forum_id":4160,"role":"guest","name":"王培松","affiliation":"中国科学院自动化研究所","sort_order":3},{"forum_id":4160,"role":"guest","name":"鲁金铭","affiliation":"南京大学","sort_order":4},{"forum_id":4160,"role":"guest","name":"冷静文","affiliation":"上海交通大学","sort_order":5},{"forum_id":4160,"role":"guest","name":"秦浩桐","affiliation":"北京航空航天大学","sort_order":6},{"forum_id":4160,"role":"guest","name":"李思坦","affiliation":"华为技术有限公司","sort_order":7}],"host":"","image_url":"/api/forums/4160/cover"},{"id":4167,"title":"Special Forum on Intelligent Computing Interconnect Technology Innovation","topic":"Unclassified","kind":"专题论坛","chair":"李论","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-408","description":"Evolution trends of intelligent computing interconnect technology and related standards progress;","is_demo":0,"cover_url":"/api/forums/4167/cover?size=thumb","cover_alt":"Special Forum on Intelligent Computing Interconnect Technology Innovation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4167","source_checked_at":"2026-10-30","agenda_json":"[[\"Insights into AI Software and Hardware Trends and Sharing of the AISHPerf Benchmark System\",\"李论\"],[\"Exploration and Practice of OISA Interconnect Technology for Next-Generation Intelligent Computing Super Nodes\",\"李锴\"],[\"Industry Trends and Innovative Practices in Intelligent Computing Interconnect\",\"李和松\"],[\"Evolution of the Communication Wall from HPC to AI\",\"何万青\"]]","code":"CNCC2026-4167","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-408","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4167","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智算互联技术创新专题论坛","original_topic":"未分类","original_description":"智算互联技术演进趋势及相关标准进展；","people":[{"forum_id":4167,"role":"chair","name":"李论","affiliation":"中国信息通信研究院","sort_order":0},{"forum_id":4167,"role":"guest","name":"李锴","affiliation":"中国移动通信有限公司研究院","sort_order":0},{"forum_id":4167,"role":"guest","name":"李和松","affiliation":"中兴通讯股份有限公司","sort_order":1},{"forum_id":4167,"role":"guest","name":"何万青","affiliation":"清微智能科技股份有限公司","sort_order":2}],"host":"","image_url":"/api/forums/4167/cover"},{"id":4169,"title":"Frontiers of Generative AI","topic":"AI Foundation Models","kind":"专题论坛","chair":"李文","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-512","description":"Generative AI is one of the most vibrant and influential frontier directions in computer science today. Advanced generative models such as large language models and diffusion models continue to achieve key technical breakthroughs, profoundly reshaping scientific research paradigms and industrial application forms. It combines deep academic research value with broad deployment potential, making it a frontier topic of great discussion value at present.\nThis symposium focuses on frontier academic progress in generative AI, conducting in-depth exchanges on the technical evolution of mainstream generative models, core scientific bottlenecks, cross-domain integration and innovation, and security and ethical norms, jointly exploring key issues in the field and future development paths.\nThe forum will invite experts, scholars, and young research talents in the field to share their latest results, build an open and inclusive academic exchange platform, promote intellectual collision and research collaboration, and drive high-quality innovative development of the domestic generative AI field.","is_demo":0,"cover_url":"/api/forums/4169/cover?size=thumb","cover_alt":"Frontiers of Generative AI","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4169","source_checked_at":"2026-10-30","agenda_json":"[[\"Empowering multi-modal large language models with active perception using reinforcement learning\",\"沈春华\"],[\"High-Quality Complex Image Generation\",\"左旺孟\"],[\"View- and Content-Controllable Video Generation\",\"郭裕兰\"],[\"Cross-Modal Content Editing\",\"赵洲\"]]","code":"CNCC2026-4169","forum_type":"专题论坛","format":null,"theme_code":"AI基础模型","theme_name":"AI基础模型","venue":"西博城5号馆","room":"西博城5号馆-512","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4169","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"生成式人工智能前沿探讨","original_topic":"AI基础模型","original_description":"生成式人工智能是当前计算机领域最具活力与影响力的前沿方向之一，大语言模型、扩散模型等先进生成模型持续实现关键技术突破，深刻重塑科研创新范式与产业应用形态，兼具深厚的学术研究价值与广阔的落地发展潜力，是当下极具研讨价值的前沿选题。\n本次研讨会聚焦生成式人工智能前沿学术进展，围绕主流生成模型的技术演进脉络、核心科学瓶颈、跨领域融合创新、安全伦理规范等方向展开深度交流，共同探讨领域关键问题与未来发展路径。\n论坛将邀请领域专家学者与青年科研骨干分享最新成果，搭建开放包容的学术交流平台，促进思想碰撞与科研协作，推动国内生成式人工智能领域高质量创新发展。","people":[{"forum_id":4169,"role":"guest","name":"沈春华","affiliation":"浙江大学","sort_order":0},{"forum_id":4169,"role":"chair","name":"李文","affiliation":"电子科技大学","sort_order":0},{"forum_id":4169,"role":"guest","name":"左旺孟","affiliation":"哈尔滨工业大学","sort_order":1},{"forum_id":4169,"role":"guest","name":"郭裕兰","affiliation":"中山大学","sort_order":2},{"forum_id":4169,"role":"guest","name":"赵洲","affiliation":"浙江大学 计算机学院","sort_order":3},{"forum_id":4169,"role":"guest","name":"刘昭强","affiliation":"电子科技大学","sort_order":4}],"host":"","image_url":"/api/forums/4169/cover"},{"id":4175,"title":"Redefining Trust Boundaries: Synergy Between Intrinsic Security and Peripheral Defenses for Large Models","topic":"Unclassified","kind":"专题论坛","chair":"吴涛、冉靖","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-516","description":"Currently, large model security is at a critical stage where capability development and security governance proceed in parallel. On the one hand, the security problems of large models themselves, such as hallucinations, jailbreaking, and alignment deviations, have not yet been fundamentally resolved and remain a bottom line that must be upheld. On the other hand, large models are evolving from content generation to driving actions, supporting AI Agents in autonomous planning, tool invocation, and task execution, and the security boundary is accordingly expanding from inside the model to multiple links such as planning, execution, and interaction. This brings new challenges: security boundaries are becoming blurred, and the focus of protection is no longer limited to model outputs; there is still a lack of mature solutions for how intrinsic security and peripheral defenses should collaborate; and traditional content moderation methods can hardly effectively constrain AI Agent behavior. In practice, defense costs are high, risk identification is difficult, and cross-link protection is weak, so existing defense approaches can hardly adapt to development needs. How to shift from passive response and single-point protection to proactive prevention and full-chain assurance has become an important issue urgently in need of discussion.\nThis perspective forum will hold discussions around issues such as the security boundaries of large models, the collaboration between intrinsic security and peripheral guardrails, and the shift from content security to behavior security, promote deep integration of academia and industry, assess the development direction of next-generation artificial intelligence security defense systems, clarify core capabilities, overall architecture, and implementation paths, and form a construction roadmap that balances practical feasibility and future adaptability.\nDebate point 1: What is the relationship between large model security and AI Agent security? How do their security boundaries, risk characteristics, and protection priorities differ?\nDebate point 2: How can a large model security ecosystem be built? Between intrinsic security and peripheral guardrails, which is more important?\nDebate point 3: From \"managing content\" to \"managing behavior\": under large model security conditions, how can implementable AI Agent security be achieved?","is_demo":0,"cover_url":"/api/forums/4175/cover?size=thumb","cover_alt":"Redefining Trust Boundaries: Synergy Between Intrinsic Security and Peripheral Defenses for Large Models","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4175","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4175","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-516","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":0,"forum_id":"CNCC2026-4175","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"重构可信边界：大模型内生安全与外围防线的协同之道","original_topic":"未分类","original_description":"当前，大模型安全正处于能力发展与安全治理并行的关键阶段。一方面，大模型自身的安全问题，如幻觉、越狱、对齐偏差等，尚未得到根本解决，仍是必须坚守的底线。另一方面，大模型正从内容生成向驱动行动演进，支撑智能体进行自主规划、工具调用与任务执行，安全边界随之从模型内部扩展至规划、执行、交互等多个环节。由此带来新的挑战：安全边界趋于模糊，防护重点不再局限于模型输出；内生安全与外围防线如何协同，尚缺乏成熟方案；传统内容审核手段难以有效约束智能体行为。实践中，防御成本高、风险识别难、跨环节防护薄弱，既有防御思路难以适应发展需求。如何从被动应对、单点防护转向主动预防、全链路保障，已成为亟待探讨的重要议题。\n本次观点论坛将围绕大模型安全边界、内生安全与外围护栏协同、从内容安全到行为安全等问题展开研讨，推动学术与产业深度融合，研判下一代人工智能安全防御体系的发展方向，明确核心能力、总体架构与实施路径，形成兼顾现实可行性与未来适应性的建设路线图。\n思辨点一：大模型安全与智能体安全是什么关系？二者的安全边界、风险特征与防护重点有何不同？\n思辨点二：如何构建大模型安全生态？内生安全与外围护栏，孰轻孰重？\n思辨点三：从“管内容”到“管行为”：大模型安全条件下如何实现可落地的智能体安全？","people":[{"forum_id":4175,"role":"chair","name":"吴涛","affiliation":"重庆邮电大学","sort_order":0},{"forum_id":4175,"role":"guest","name":"张清华","affiliation":"重庆邮电大学","sort_order":0},{"forum_id":4175,"role":"co_chair","name":"冉靖","affiliation":"绿盟科技集团股份有限公司","sort_order":1},{"forum_id":4175,"role":"guest","name":"赵翔","affiliation":"国防科技大学","sort_order":1},{"forum_id":4175,"role":"guest","name":"贺剑","affiliation":"重庆长安汽车股份有限公司","sort_order":2},{"forum_id":4175,"role":"guest","name":"李华雄","affiliation":"南京大学","sort_order":3},{"forum_id":4175,"role":"guest","name":"牛伟纳","affiliation":"电子科技大学","sort_order":4},{"forum_id":4175,"role":"guest","name":"向国勇","affiliation":"绿盟科技集团股份有限公司","sort_order":5}],"host":"","image_url":"/api/forums/4175/cover"},{"id":4186,"title":"Experiencing Barriers, Reconstructing Accessibility: Building an Inclusive and Equitable Digital Research Ecosystem","topic":"General","kind":"专题论坛","chair":"王甦菁","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-104黄龙","description":"Researchers with disabilities in computer science generally face unique real-world barriers throughout the entire process of doctoral study, experimentation, paper publication, and job seeking, such as lack of device accessibility, high thresholds for accessing academic resources, social stereotypes, and sustained physical and mental burdens. To break through research obstacles, they proactively study digital technologies and are technological explorers who continuously explore the path of information accessibility; moreover, based on the perspective of their own real barriers, they conduct original algorithm research that in turn optimizes accessibility solutions such as intelligent assistive devices, human-computer interaction, and multimodal large models, becoming core co-builders of the AI accessibility ecosystem. This forum creates a dedicated communication platform, focusing on the real pain points throughout the full-cycle research of this group, sharing replicable practical experience in academic perseverance, research innovation, and career development, opening channels for mutual exchange among young researchers with disabilities across universities, uncovering the unique accessibility innovation value of scholars with disabilities, providing practical evidence for research institutions to improve supporting support for talent with disabilities, helping more students with disabilities in science and engineering devote themselves to computer science research, and reconstructing an inclusive and equitable digital research ecosystem through firsthand practice.","is_demo":0,"cover_url":"/api/forums/4186/cover?size=thumb","cover_alt":"Experiencing Barriers, Reconstructing Accessibility: Building an Inclusive and Equitable Digital Research Ecosystem","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4186","source_checked_at":"2026-10-30","agenda_json":"[[\"From Humanistic Care to Technological Innovation: AI Bias and Inclusive AI Empowerment from an Interdisciplinary Perspective\",\"矣晓沅\"],[\"Making Intelligence an Extension of Me: From Researching Large Models to Doing Research with Large Models\",\"邵镇炜\"],[\"From Code to Accessibility: A Disabled Computer Researcher's Exploration and Practice of AI for Disability Assistance\",\"林家钦\"],[\"From Needs to Innovation: A Disabled PhD Student's Exploration of Artificial Intelligence Research\",\"康新晨\"],[\"From Support to Empowerment: A Disabled Researcher's Educational Practice and Exploration of Information Accessibility\",\"姚登峰\"]]","code":"CNCC2026-4186","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"天府国际会议中心","room":"天府国际会议中心-104黄龙","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4186","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"亲历障碍，重构无碍：共建包容平等的数字科研生态","original_topic":"综合","original_description":"残障计算机研究者在读博、实验、论文发表、求职全过程中，普遍面临设备适配缺失、学术资源获取门槛高、社会刻板认知、持续身心负担等独有现实障碍。为突破科研阻碍，他们主动钻研数字技术，是信息无障碍道路上持续探索的科技求索者；更立足自身真实障碍视角开展原创算法研究，反向优化智能辅具、人机交互、多模态大模型等无障碍方案，成为AI无障碍生态的核心共建者。本论坛搭建专属交流载体，聚焦该群体全周期科研真实痛点，分享可复制的求学攻坚、科研创新、职业发展实操经验，打通跨校残障青年科研人才互通渠道，挖掘残障学者独有的无障碍创新价值，为科研单位完善残障人才配套支持提供实践依据，助力更多残障理工学子投身计算机科研，以亲身实践重构包容平等的数字科研生态。","people":[{"forum_id":4186,"role":"chair","name":"王甦菁","affiliation":"中国科学院心理研究所","sort_order":0},{"forum_id":4186,"role":"guest","name":"袁鑫","affiliation":"武汉科技大学","sort_order":0},{"forum_id":4186,"role":"guest","name":"矣晓沅","affiliation":"微软亚洲研究院","sort_order":1},{"forum_id":4186,"role":"guest","name":"邵镇炜","affiliation":"杭州电子科技大学","sort_order":2},{"forum_id":4186,"role":"guest","name":"林家钦","affiliation":"上海大学","sort_order":3},{"forum_id":4186,"role":"guest","name":"康新晨","affiliation":"南京理工大学","sort_order":4},{"forum_id":4186,"role":"guest","name":"吴竞琦","affiliation":"南方科技大学，北京中关村学院","sort_order":5},{"forum_id":4186,"role":"guest","name":"王戟","affiliation":"国防科技大学计算机学院","sort_order":6},{"forum_id":4186,"role":"guest","name":"姚登峰","affiliation":"北京联合大学","sort_order":7}],"host":"","image_url":"/api/forums/4186/cover"},{"id":4188,"title":"The Next Stop for Financial Agents: Who Decides, Who Executes, Who Is Accountable?","topic":"Unclassified","kind":"专题论坛","chair":"吕智慧、敖翔","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-504","description":"This forum, in the form of deliberative dialogue, discusses how technical capabilities and action authority should be matched when financial agents move from information analysis toward autonomous research, decision-making, and execution. The forum will provide guiding reports from four directions: Can the representation, learning, and dynamic evolution of big data association structures help agents understand complex relationships and predict changes? Can blockchain architectures, payment mechanisms, and transaction supervision for the agent economy support autonomous collaboration and leave verifiable execution records? How should the professional understanding, logical reasoning, and safety capabilities of financial large models be evaluated? And how can financial deep research agents organize evidence, complete research tasks, and prove that their conclusions are trustworthy?\nAfter the reports, the forum will further discuss: Can agents with good evaluation performance obtain decision-making and execution authority in real environments? Can on-chain records prove \"what was done,\" and can they explain \"why it was done this way\" and support responsibility attribution? When associated data, model reasoning, and agent research conclusions conflict with one another, whom should the system believe, and who should review? Through cross-discussion of data, models, architectures, and execution chains, the forum will explore controllable, trustworthy, and accountable technical paths for financial agents.","is_demo":0,"cover_url":"/api/forums/4188/cover?size=thumb","cover_alt":"The Next Stop for Financial Agents: Who Decides, Who Executes, Who Is Accountable?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4188","source_checked_at":"2026-10-30","agenda_json":"[[\"New Challenges for Blockchain Technology in the Agent Economy\",\"成秀珍\"],[\"Large-Model Agents for Financial Scenarios and Their Evaluation Research\",\"程大伟\"],[\"Computational Models of Association Structures in Big Data and Their Applications\",\"李建欣\"],[\"Toward Professional and Trustworthy Financial Deep Research Agents: Evaluation, Techniques, and Outlook\",\"朱凤彬\"],[\"FinStyle: Controllable Delivery of High-Information-Density Content Creation Agents in Financial Scenarios\",\"刘继洋\"]]","code":"CNCC2026-4188","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-504","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4188","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"金融智能体的下一站：谁来决策，谁来执行，谁来担责？","original_topic":"未分类","original_description":"本论坛以思辨对话的形式，讨论金融智能体从信息分析走向自主研究、决策与执行时，技术能力与行动权限应如何匹配。论坛将从四个方向进行引导报告：大数据关联结构的表征、学习与动态演化，能否帮助智能体理解复杂关系并预测变化？面向智能体经济的区块链架构、支付机制与交易监管，能否支撑自主协作并留下可验证的执行记录？金融大模型的专业理解、逻辑推理与安全能力，应如何评测？金融深度研究智能体又如何组织证据、完成研究任务，并证明其结论可信？\n报告之后，论坛将进一步研讨：评测表现良好的智能体，是否可以获得真实环境中的决策权和执行权？链上记录能够证明“做过什么”，能否解释“为什么这样做”并支持责任归因？当关联数据、模型推理和智能体研究结论彼此冲突时，系统应相信谁、由谁复核？论坛将通过对数据、模型、架构与执行链条的交叉讨论，探索可控、可信、可追责的金融智能体技术路径。","people":[{"forum_id":4188,"role":"chair","name":"吕智慧","affiliation":"复旦大学","sort_order":0},{"forum_id":4188,"role":"guest","name":"成秀珍","affiliation":"山东大学","sort_order":0},{"forum_id":4188,"role":"co_chair","name":"敖翔","affiliation":"中国科学院计算所","sort_order":1},{"forum_id":4188,"role":"guest","name":"程大伟","affiliation":"同济大学","sort_order":1},{"forum_id":4188,"role":"guest","name":"李建欣","affiliation":"北京航空航天大学","sort_order":2},{"forum_id":4188,"role":"guest","name":"朱凤彬","affiliation":"新加坡国立大学","sort_order":3},{"forum_id":4188,"role":"guest","name":"刘继洋","affiliation":"蚂蚁集团","sort_order":4},{"forum_id":4188,"role":"guest","name":"杨新","affiliation":"西南财经大学","sort_order":5}],"host":"","image_url":"/api/forums/4188/cover"},{"id":4189,"title":"The 4th Confidential Computing Forum: Integration and Deployment of Confidential Computing with the AI Industry","topic":"AI Safety and Governance","kind":"专题论坛","chair":"冯登国、张锋巍","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-405","description":"The forum will mainly discuss the value and impact of confidential computing on the industrial AI industry. It will focus on the important issue of balancing security, performance, and overhead in confidential computing-assisted AI computation. At present, AI applications such as large models have gradually become a trend in the internet industry, and the industry is also gradually paying attention to the importance of AI to its industrial development (for example, information processing, code programming, and intelligent interaction). However, in this process, problems such as leakage of confidential information are inevitably present. Therefore, this forum takes \"integration of confidential computing and the AI industry\" as its topic, a practical-scenario-oriented topic with broad impact, and will focus on discussing the practical issue of combining confidential computing, an important technology in the field of system security, with AI industries such as large language models, the importance of confidential computing for tokens and privacy protection in large language models, and the difficulties and challenges of actually deploying confidential computing on enterprise cloud platforms, so as to provide an important reference for the development of data security in the domestic internet industry.","is_demo":0,"cover_url":"/api/forums/4189/cover?size=thumb","cover_alt":"The 4th Confidential Computing Forum: Integration and Deployment of Confidential Computing with the AI Industry","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4189","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Remarks\",\"冯登国\"],[\"Zero-Trust-Oriented Confidential AI Technologies and Systems\",\"秦宇\"],[\"Research on Confidential Computing Technologies for Large Model Inference\",\"丁滟\"],[\"Formal Verification Technologies for Trusted Execution Environments Oriented to the GPTEE Standard\",\"赵永望\"],[\"Confidential Computing Cluster Framework Technologies and Applications for Large Models\",\"张磊\"],[\"Confidential Device Drivers\",\"郭力维\"],[\"Discussion Session --- Industry Deployment Panel Moderator\",\"侯锐\"],[\"Panel Discussion---Industry Deployment Panel\",\"郭御风\"],[\"Panel Discussion---Industry Deployment Panel\",\"闫守孟\"],[\"Panel Discussion---Industry Deployment Panel\",\"吴烨\"],[\"Panel Discussion---Industry Deployment Panel\",\"金意儿\"],[\"Panel Discussion---Industry Deployment Panel\",\"应志伟\"]]","code":"CNCC2026-4189","forum_type":"专题论坛","format":null,"theme_code":"AI安全与治理","theme_name":"AI安全与治理","venue":"西博城4号馆","room":"西博城4号馆-405","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":12,"forum_id":"CNCC2026-4189","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第四届机密计算论坛——机密计算与AI产业的结合与落地","original_topic":"AI安全与治理","original_description":"论坛将主要研讨机密计算对工业界AI产业的价值与影响。着眼于机密计算辅助AI计算的安全-性能-开销平衡这一重要问题。当前，大模型等AI应用逐渐成为互联网产业的潮流，工业界也逐渐注重AI对其产业发展（例如信息处理、代码编程、智能交互）的重要性，然而在该过程中，难免存在机密信息泄露等问题。因此，本论坛将“机密计算与AI产业结合”这一面向实际场景的、影响广泛的作为论坛选题，将着重讨论机密计算这一系统安全领域重要技术与大语言模型等AI产业结合的现实问题，机密计算对大语言模型中token、隐私保护的重要性，将机密计算实际落地部署于企业云平台的难点与挑战，并为国内互联网产业数据安全发展提供重要参考。","people":[{"forum_id":4189,"role":"chair","name":"冯登国","affiliation":"中国科学院软件研究所","sort_order":0},{"forum_id":4189,"role":"guest","name":"秦宇","affiliation":"中科院软件研究所","sort_order":0},{"forum_id":4189,"role":"co_chair","name":"张锋巍","affiliation":"南方科技大学","sort_order":1},{"forum_id":4189,"role":"guest","name":"丁滟","affiliation":"国防科技大学","sort_order":1},{"forum_id":4189,"role":"guest","name":"赵永望","affiliation":"浙江大学","sort_order":2},{"forum_id":4189,"role":"guest","name":"张磊","affiliation":"南湖实验室","sort_order":3},{"forum_id":4189,"role":"guest","name":"郭力维","affiliation":"电子科技大学","sort_order":4},{"forum_id":4189,"role":"guest","name":"侯锐","affiliation":"中国科学院信息工程研究所","sort_order":5},{"forum_id":4189,"role":"guest","name":"郭御风","affiliation":"飞腾信息技术有限公司","sort_order":6},{"forum_id":4189,"role":"guest","name":"闫守孟","affiliation":"蚂蚁集团","sort_order":7},{"forum_id":4189,"role":"guest","name":"吴烨","affiliation":"字节跳动","sort_order":8},{"forum_id":4189,"role":"guest","name":"金意儿","affiliation":"华为技术有限公司","sort_order":9},{"forum_id":4189,"role":"guest","name":"应志伟","affiliation":"海光信息技术股份有限公司","sort_order":10}],"host":"","image_url":"/api/forums/4189/cover"},{"id":4198,"title":"Intelligence for Life: AI Empowering Life and Health Frontier Forum","topic":"AI for Industry","kind":"专题论坛","chair":"苏苒、裴红斌","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-413","description":"The deep cross-disciplinary integration of artificial intelligence and the life and health field is driving innovation across the entire chain of biomedical research, from data acquisition and intelligent analysis to scientific discovery and clinical application. This forum focuses on frontier directions at the deep intersection of AI and biomedicine, bringing together hot topics such as ultrasonic microscopy imaging, intelligent drug design, biomedical big data computing, multi-omics analysis, and multimodal intelligent analysis, and will exchange views on key scientific questions and core technologies in the life and health field, from precise perception and complex data understanding to intelligent discovery. The forum will showcase the latest advances of artificial intelligence in medical imaging, drug development, omics data mining, and intelligent analysis in data-scarce scenarios, and explore how AI can accelerate the discovery of life principles and biomedical innovation, opening up new ideas, new methods, and new opportunities for the deep integration of intelligent technologies and life sciences.","is_demo":0,"cover_url":"/api/forums/4198/cover?size=thumb","cover_alt":"Intelligence for Life: AI Empowering Life and Health Frontier Forum","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4198","source_checked_at":"2026-10-30","agenda_json":"[[\"Building Precision Oncology Treatment Models Based on Microbiomics\",\"苏苒\"],[\"Intelligent Computing Models for Biomedical Big Data: Algorithm Design and Practice\",\"陈慧灵\"],[\"Analysis and Understanding of Multi-Omics Data\",\"李向涛\"],[\"Research Progress in AI-Based Drug Design\",\"谢昌谕\"],[\"Intelligent Analysis Methods for Multimodal Data-Constrained Biomedical Scenarios\",\"金强国\"]]","code":"CNCC2026-4198","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"西博城4号馆","room":"西博城4号馆-413","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4198","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智汇生命 · 人工智能赋能生命健康前沿论坛","original_topic":"AI+行业","original_description":"人工智能与生命健康领域的深度交叉融合，正在推动生物医学研究从数据获取、智能分析到科学发现与临床应用的全链条创新。本论坛聚焦AI与生物医学深度交叉的前沿方向，汇集超声显微成像、智能药物设计、生物医学大数据计算、多组学解析及多模态智能分析等热点议题，围绕生命健康领域从精准感知、复杂数据理解到智能发现的关键科学问题与核心技术展开交流。论坛将展示人工智能在医学成像、药物研发、组学数据挖掘和数据稀缺场景智能分析中的最新进展，探讨AI如何加速生命规律发现与生物医学创新，为智能技术与生命科学的深度融合开拓新思路、新方法与新机遇。","people":[{"forum_id":4198,"role":"chair","name":"苏苒","affiliation":"天津大学","sort_order":0},{"forum_id":4198,"role":"guest","name":"陈慧灵","affiliation":"温州大学","sort_order":0},{"forum_id":4198,"role":"co_chair","name":"裴红斌","affiliation":"西安交通大学","sort_order":1},{"forum_id":4198,"role":"guest","name":"李向涛","affiliation":"吉林大学","sort_order":1},{"forum_id":4198,"role":"guest","name":"谢昌谕","affiliation":"浙江大学","sort_order":2},{"forum_id":4198,"role":"guest","name":"金强国","affiliation":"西北工业大学","sort_order":3}],"host":"","image_url":"/api/forums/4198/cover"},{"id":4201,"title":"[Viewpoint Forum] AI Reshapes Production Relations: A Harsh Winter of Layoffs, or Organizational Rebirth?","topic":"General","kind":"专题论坛","chair":"陈昊、汪卫国","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-118A峨眉厅A","description":"Artificial intelligence is moving from an auxiliary tool to a production subject. Large models and AI Agents are deeply embedded in the entire enterprise chain, directly driving decision-making and execution, and reshaping production methods and organizational logic. Throughout history, leaps in productivity have always been accompanied by organizational change: clans in the agricultural era, bureaucracy in the industrial era, and platformization in the information era. Today, while AI replaces repetitive work, it reconstructs division of labor and collaboration, and the value of interdisciplinary talent with system design capabilities becomes prominent.\nAt present, social consensus has not yet formed: job compression coexists with high-salary opportunities, and layoff anxiety coexists with career dividends. This is not a simple employment fluctuation, but a structural pain of production relations. Enterprises urgently need to transform, universities urgently need to reconstruct training systems, and individuals need to find irreplaceable value.\nDiscussion topics:\n1 Is AI layoffs an efficiency revolution, or organizational imbalance?\n2 After AI reshapes production relations, how much human value remains?\n3 Crossing the harsh winter of the era, how should talent growth be reconstructed?","is_demo":0,"cover_url":"/api/forums/4201/cover?size=thumb","cover_alt":"[Viewpoint Forum] AI Reshapes Production Relations: A Harsh Winter of Layoffs, or Organizational Rebirth?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4201","source_checked_at":"2026-10-30","agenda_json":"[[\"AI Reshaping Production Relations: Universities' Value Persistence and Educational Transformation\",\"翟东海\"],[\"In the AI Era, How Investment Redefines the Value of \\\"People\\\"\",\"张钟月\"],[\"Exploration and Practice of Corporate \\\"Talent Strategy\\\" in the AI Era\",\"刘赞\"]]","code":"CNCC2026-4201","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"天府国际会议中心","room":"天府国际会议中心-118A峨眉厅A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":3,"forum_id":"CNCC2026-4201","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"【观点论坛】AI重塑生产关系：裁员凛冬，还是组织新生？","original_topic":"综合","original_description":"人工智能正从辅助工具迈向生产主体。大模型、智能体深度嵌入企业全链条，直接驱动决策与执行，重塑生产方式与组织逻辑。纵观历史，生产力跃迁必伴生组织变革：农业时代宗族、工业时代科层制、信息时代平台化。如今，AI在替代重复性工作的同时，重构分工与协作，具备系统设计能力的复合型人才价值凸显。\n当前社会共识尚未形成：岗位压缩与高薪机会并存，裁员焦虑与职业红利同在。这并非简单就业波动，而是生产关系的结构性阵痛。企业亟待转型，高校亟需重构培养体系，个体更需寻找不可替代的价值。\n思辨议题：\n1AI 裁员是效率革命，还是组织失衡？\n2AI 重塑生产关系后，人的价值还剩多少？\n3跨越时代的凛冬，人才成长应如何重构？","people":[{"forum_id":4201,"role":"chair","name":"陈昊","affiliation":"微软亚洲研究院","sort_order":0},{"forum_id":4201,"role":"guest","name":"翟东海","affiliation":"西南交通大学","sort_order":0},{"forum_id":4201,"role":"co_chair","name":"汪卫国","affiliation":"新华三技术有限公司","sort_order":1},{"forum_id":4201,"role":"guest","name":"张钟月","affiliation":"北极光创投","sort_order":1},{"forum_id":4201,"role":"guest","name":"刘赞","affiliation":"新华三技术有限公司","sort_order":2},{"forum_id":4201,"role":"guest","name":"张津","affiliation":"微软亚洲研究院","sort_order":3},{"forum_id":4201,"role":"guest","name":"熊程","affiliation":"成都大学","sort_order":4},{"forum_id":4201,"role":"guest","name":"刘伟","affiliation":"小米","sort_order":5}],"host":"","image_url":"/api/forums/4201/cover"},{"id":4209,"title":"AI Ready: Are AI Agents Ready to \"Take the Job\"?","topic":"Unclassified","kind":"专题论坛","chair":"陈小军、章秦","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-513","description":"Large models are moving from \"generating content\" to \"executing tasks,\" and AI agents are evolving from auxiliary tools into action entities capable of autonomous planning, tool invocation, and closed-loop execution. Digital agents are beginning to penetrate enterprise systems and business processes; embodied agents such as robots and drones are entering the physical world to perform perception, decision-making, and manipulation; and scientific agents are further participating in hypothesis generation, experiment design, and autonomous exploration. The integration of digital intelligence and physical intelligence makes it possible for AI to span digital systems and the physical world and continuously complete complex tasks.\nAs AI moves from \"assisting people\" to \"undertaking tasks,\" a practical question arises: is AI truly Ready to \"take the job\"? Here, Ready does not mean that a model performs strongly enough on benchmarks, nor that an agent can complete a few demonstration tasks, but whether it can be competent for the corresponding work in real scenarios, form effective collaboration with people and other agents, and remain trustworthy and dependable after gaining greater autonomy.\nThis forum discusses three progressively layered questions: capability boundaries, organizational structures, and trust assurance. First, as tasks extend from digital space to the physical world and from deterministic processes to open scientific exploration, how can we further characterize an agent's \"job competency\" for real work beyond the model's individual capabilities? Second, as digital agents, robots, and others begin to become new task-execution entities, how will the traditional way of working centered on \"people—positions—processes\" evolve, and what new division of labor and collaboration relationships will form among people, Agents, and machines? Further, when agents can autonomously plan, decide, and take real actions, how can we establish trust mechanisms matching their capabilities and degree of autonomy, so that they truly move from \"able to execute\" to \"can be entrusted\"?\nThe forum will focus on typical scenarios such as enterprise operations, production services, scientific research, and physical manipulation, inviting experts from academia and industry to jointly examine the key transitions of agents from model capabilities to job competency, from auxiliary tools to collaborative entities, and from usable to trustworthy, and to further explore: when agents truly begin to \"take the job,\" are our capability evaluation, organizational methods, and trust systems also ready?","is_demo":0,"cover_url":"/api/forums/4209/cover?size=thumb","cover_alt":"AI Ready: Are AI Agents Ready to \"Take the Job\"?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4209","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4209","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-513","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":0,"forum_id":"CNCC2026-4209","legacy_id":5,"lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI Ready：智能体准备好“上岗”了吗？","original_topic":"未分类","original_description":"大模型正从“生成内容”走向“执行任务”，智能体也从辅助工具演进为能够自主规划、调用工具并完成闭环执行的行动主体。数字智能体开始深入企业系统与业务流程，机器人、无人机等具身智能体进入物理世界完成感知、决策与操作，科学智能体则进一步参与假设生成、实验设计与自主探索。数字智能与物理智能的融合，使 AI 开始具备跨越数字系统与物理世界、持续完成复杂任务的可能。\n当 AI 从“辅助人”走向“承担任务”，一个现实问题随之而来：AI 是否真的 Ready，可以“上岗”了？ 这里的 Ready，不是模型在 Benchmark 上表现得足够强，也不是智能体能够完成若干演示任务，而是它能否在真实场景中胜任相应工作，与人及其他智能体形成有效协作，并在获得更大自主权后仍然值得信任和托付。\n本论坛围绕能力边界、组织架构与可信保障三个层层递进的问题展开讨论。首先，当任务从数字空间延伸到物理世界、从确定性流程走向开放科学探索，如何从模型的单项能力进一步刻画智能体面向真实工作的“岗位胜任力”？其次，当数字智能体、机器人等开始成为新的任务执行主体，传统以“人—岗位—流程”为核心的工作方式将如何演化，人、Agent与机器又将形成怎样的新型分工与协作关系？进一步，当智能体能够自主规划、决策并采取真实行动，如何建立与其能力和自主程度相匹配的可信机制，使其从“能够执行”真正走向“可以托付”？\n论坛将聚焦企业经营、生产服务、科学研究及物理操作等典型场景，邀请学术界与产业界专家共同审视智能体从模型能力走向岗位胜任、从辅助工具走向协作主体、从可用走向可托付的关键跃迁，并进一步探讨：当智能体真正开始“上岗”，我们的能力评价、组织方式与可信体系是否也已经做好准备？","people":[{"forum_id":4209,"role":"chair","name":"陈小军","affiliation":"深圳大学","sort_order":0},{"forum_id":4209,"role":"guest","name":"张良杰","affiliation":"深圳大学","sort_order":0},{"forum_id":4209,"role":"co_chair","name":"章秦","affiliation":"深圳大学","sort_order":1},{"forum_id":4209,"role":"guest","name":"陈科海","affiliation":"哈尔滨工业大学（深圳）","sort_order":1},{"forum_id":4209,"role":"guest","name":"魏巍","affiliation":"山西大学","sort_order":2},{"forum_id":4209,"role":"guest","name":"Mingjiang Liang","affiliation":"深圳大学","sort_order":3},{"forum_id":4209,"role":"guest","name":"范广辉","affiliation":"京东","sort_order":4},{"forum_id":4209,"role":"guest","name":"黄超","affiliation":"香港大学","sort_order":5}],"host":"","image_url":"/api/forums/4209/cover"},{"id":4214,"title":"CCF-Ant Research Fund Closed-Door Forum: Token Efficiency Special Topic Guideline Co-Creation Meeting","topic":"Unclassified","kind":"专题论坛","chair":"邵青","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-406B兴隆湖B","description":"This is a closed-door forum focused on joint academic and industrial exploration of Token Efficiency. Centered on fund topics as the core vehicle, this forum specially invites academic experts and Ant industrial experts to engage in in-depth closed-door dialogue. The forum aims to break down the cognitive boundaries between academia and industry and, through a model of \"topic-driven, expert-led, enterprise-collaborative,\" focuses on a panoramic discussion of the future development plans and expected outcome transformation of the topics. We will deeply analyze technical difficulties and look ahead to industrial layout, aiming through high-density intellectual exchange to make fund topics a solid bridge connecting scientific research dreams with industrial reality and to jointly pursue a new future of industry-academia-research cooperation.","is_demo":0,"cover_url":"/api/forums/4214/cover?size=thumb","cover_alt":"CCF-Ant Research Fund Closed-Door Forum: Token Efficiency Special Topic Guideline Co-Creation Meeting","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4214","source_checked_at":"2026-10-30","agenda_json":"[[\"Conference Remarks\",\"骆骥\"],[\"Ant's Practices and Reflections on Token Efficiency\",\"黄挺\"],[\"Prediction-Based Scheduling for Large Model Inference Systems\",\"杨波威\"]]","code":"CNCC2026-4214","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-406B兴隆湖B","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":3,"forum_id":"CNCC2026-4214","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"CCF-蚂蚁科研基金闭门论坛-Token Efficiency专项课题指南共创会","original_topic":"未分类","original_description":"这是一场聚焦学术与产业的共同探讨Token Efficiency的闭门论坛。本次论坛以基金课题为核心载体，特邀学术专家和蚂蚁产业专家，展开深度闭门对话。论坛旨在打破学界与业界的认知边界，通过“课题牵引、专家主导、企业协同”的模式，重点围绕课题的未来发展规划与预期成果转化进行全景式探讨。我们将深入剖析技术难点，前瞻产业布局，旨在通过高浓度的智慧激荡，让基金课题成为连接科研梦想与产业现实的坚实桥梁，共谋产学研合作新未来。","people":[{"forum_id":4214,"role":"chair","name":"邵青","affiliation":"蚂蚁集团","sort_order":0},{"forum_id":4214,"role":"guest","name":"骆骥","affiliation":"蚂蚁集团","sort_order":0},{"forum_id":4214,"role":"guest","name":"金海","affiliation":"华中科技大学","sort_order":1},{"forum_id":4214,"role":"guest","name":"过敏意","affiliation":"上海交通大学","sort_order":2},{"forum_id":4214,"role":"guest","name":"彭涛","affiliation":"蚂蚁集团","sort_order":3},{"forum_id":4214,"role":"guest","name":"黄挺","affiliation":"蚂蚁集团","sort_order":4},{"forum_id":4214,"role":"guest","name":"漆桂林","affiliation":"东南大学","sort_order":5},{"forum_id":4214,"role":"guest","name":"顾荣","affiliation":"南京大学","sort_order":6},{"forum_id":4214,"role":"guest","name":"李诚","affiliation":"中国科学技术大学","sort_order":7},{"forum_id":4214,"role":"guest","name":"杨波威","affiliation":"浙江大学航空航天学院","sort_order":8}],"host":"","image_url":"/api/forums/4214/cover"},{"id":4224,"title":"Can Multi-Agent Harness Build Clinically Traceable Reasoning and a New Paradigm for Medical Research?","topic":"AI4Science","kind":"专题论坛","chair":"张彤、张帆","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-501","description":"Generative AI is comprehensively reshaping the boundaries of clinical practice and research in digital medicine. Single large models commonly suffer from issues such as black-box reasoning and outputs lacking governance constraints, making it difficult to meet the requirements for compliant clinical deployment. Harness Engineering, as a standardized control foundation for agents, relies on a \"model + Harness\" architecture to achieve multi-agent collaborative scheduling, retention of reasoning chains, alignment of multimodal medical evidence, and full-process observable governance, providing a new path for building trustworthy medical intelligence systems.\nThis forum invites multiple experts in the field of digital medicine to introduce frontiers of multi-agent clinical research and engage in in-depth deliberation, jointly discussing how medical multi-agent systems can drive digital medicine toward standardization and build a fully traceable intelligent reasoning system.","is_demo":0,"cover_url":"/api/forums/4224/cover?size=thumb","cover_alt":"Can Multi-Agent Harness Build Clinically Traceable Reasoning and a New Paradigm for Medical Research?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4224","source_checked_at":"2026-10-30","agenda_json":"[[\"Efficient Training Strategies for Deep Learning Models: Model Training in the Era of LLMs and AI Agents\",\"宋志坚\"],[\"Multi-Agent-Based AI-Assisted Healthcare\",\"谢伟迪\"],[\"General-Purpose Intelligent Ultrasound: From Image Understanding and Brain-Inspired Knowledge Generalization to Embodied Decision-Making Systems\",\"黄庆华\"],[\"Research on an Evidence-Driven Trustworthy Multi-Agent Reasoning Architecture for Renal Transplant Pathology\",\"徐增林\"],[\"Development and Application of Multi-Agent Systems for Cardiovascular Disease\",\"何怡华\"],[\"Roundtable Discussion\",\"吴健\"],[\"Roundtable Discussion\",\"杨士强\"]]","code":"CNCC2026-4224","forum_type":"专题论坛","format":null,"theme_code":"AI4Science","theme_name":"AI4Science","venue":"西博城5号馆","room":"西博城5号馆-501","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":7,"forum_id":"CNCC2026-4224","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"多智能体Harness能否构建临床可溯推理与医学科研新范式？","original_topic":"AI4Science","original_description":"生成式 AI 正全方位重塑数字医学临床与科研边界。单一大模型普遍存在推理黑盒、输出缺乏管控约束等问题，难以适配临床合规落地需求。驾驭工程（Harness Engineering）作为智能体标准化管控底座，依托 “模型 + Harness” 架构实现多智能体协同调度、推理链路留存、多模态医学证据对齐与全流程可观测治理，为可信医疗智能体系建设提供全新路径。\n本论坛邀请多位数字医学领域专家介绍多智能体临床研究前沿，并开展深度思辨，共议医学多智能体系统如何推动数字医学走向标准化，搭建全程可追溯智能推理体系。","people":[{"forum_id":4224,"role":"chair","name":"张彤","affiliation":"鹏城实验室","sort_order":0},{"forum_id":4224,"role":"guest","name":"宋志坚","affiliation":"复旦大学","sort_order":0},{"forum_id":4224,"role":"co_chair","name":"张帆","affiliation":"电子科技大学","sort_order":1},{"forum_id":4224,"role":"guest","name":"谢伟迪","affiliation":"上海交通大学","sort_order":1},{"forum_id":4224,"role":"guest","name":"黄庆华","affiliation":"西北工业大学","sort_order":2},{"forum_id":4224,"role":"guest","name":"徐增林","affiliation":"复旦大学","sort_order":3},{"forum_id":4224,"role":"guest","name":"何怡华","affiliation":"首都医科大学附属北京安贞医院","sort_order":4},{"forum_id":4224,"role":"guest","name":"吴健","affiliation":"浙江大学国际联合学院","sort_order":5},{"forum_id":4224,"role":"guest","name":"杨士强","affiliation":"清华大学","sort_order":6}],"host":"","image_url":"/api/forums/4224/cover"},{"id":4225,"title":"Tianfu·Huiyan—AI for Science Scientific Intelligence Sub-forum","topic":"Unclassified","kind":"专题论坛","chair":"章毅","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-115B九寨厅B","description":"This thematic forum focuses on the new paradigm of AI for Science, with exchanges around key links such as scientific foundation models, domestic computing power, scientific data, closed-loop dry and wet experiments, and agent collaboration. The forum will also hold the \"Tianfu·Huiyan\" launch ceremony, showcasing Sichuan's systematic exploration centered on scientific foundation models, supported by computing power and data experiment platforms, and driven by key discipline scenarios. Experts from research institutions, universities, and industry are invited to share frontier trends and practical progress, promoting deep collaboration among models, computing power, knowledge, and experimental facilities, and accelerating the evolution of scientific research processes from assisted analysis toward autonomous discovery, validation, and iteration.","is_demo":0,"cover_url":"/api/forums/4225/cover?size=thumb","cover_alt":"Tianfu·Huiyan—AI for Science Scientific Intelligence Sub-forum","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4225","source_checked_at":"2026-10-30","agenda_json":"[[\"From Sci-Tech Intelligence to AI for Science: Knowledge Services and Research Paradigm Shifts in the AI4S Era\",\"陈云伟\"],[\"From Brain Science to AI for Science: Brain-Inspired Spiking Neural Networks and Energy Efficiency\",\"屈鸿\"],[\"Empowering Computing Foundations, Connecting Industry Ecosystems: Domestic Intelligent Computing Driving AI Innovation and Development in Sichuan\",\"谢一明\"],[\"Closed-Loop Dry and Wet Experiments: From AI Prediction to Automated Experimental Validation\",\"王晓旭\"],[\"General-Purpose and Domain-Specific Integration Empowering a New Paradigm for Scientific Research: Frontier Trends and Future Opportunities of AI for Science\",\"马乐\"]]","code":"CNCC2026-4225","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-115B九寨厅B","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4225","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"天府·慧研—AI for Science科学智能分论坛","original_topic":"未分类","original_description":"本专题论坛聚焦人工智能赋能科学研究（AI for Science）新范式，围绕科学大模型、国产算力、科学数据、干湿实验闭环与智能体协同等关键环节展开交流。论坛同步举行“天府·慧研”发布仪式，集中展示四川以科学基础大模型为核心、算力数据实验平台为支撑、重点学科场景为牵引的体系化探索，邀请科研机构、高校与产业界专家分享前沿趋势与实践进展，推动模型、算力、知识与实验设施深度协同，加速科研流程从辅助分析向自主发现、验证和迭代演进。","people":[{"forum_id":4225,"role":"chair","name":"章毅","affiliation":"四川大学","sort_order":0},{"forum_id":4225,"role":"guest","name":"陈云伟","affiliation":"中国科学院成都文献情报中心","sort_order":0},{"forum_id":4225,"role":"guest","name":"屈鸿","affiliation":"电子科技大学","sort_order":1},{"forum_id":4225,"role":"guest","name":"谢一明","affiliation":"智算云腾（成都）科技有限公司","sort_order":2},{"forum_id":4225,"role":"guest","name":"王晓旭","affiliation":"深势科技（宜宾）有限公司","sort_order":3},{"forum_id":4225,"role":"guest","name":"马乐","affiliation":"上海人工智能实验室","sort_order":4}],"host":"","image_url":"/api/forums/4225/cover"},{"id":4226,"title":"Computational Intelligence Empowering Energy: A New Intelligent Paradigm from Data-Driven to Mechanism-Integrated","topic":"Unclassified","kind":"专题论坛","chair":"彭博","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城4号馆-415","description":"With the deep integration of artificial intelligence and the energy industry, computational intelligence is becoming a new quality productive force driving the high-quality development of the energy industry. With the theme \"Computational Intelligence Empowering Energy: A New Intelligent Paradigm from Data-Driven to Mechanism-Integrated,\" this forum focuses on the frontier challenges and opportunities in the booming period of energy intelligentization. The forum will conduct in-depth discussions on core topics such as multi-source heterogeneous energy data governance and knowledge mining, intelligent perception and cognition in complex industrial scenarios, mechanism-and-data integrated modeling based on physics-informed neural networks (PINNs) and large models, and production decision optimization. It aims to build an interdisciplinary exchange platform, bringing together experts and scholars in computer science and energy engineering to jointly explore innovative application paths of computational intelligence across the entire energy industry chain and promote the digital transformation and intelligent upgrading of the energy industry.","is_demo":0,"cover_url":"/api/forums/4226/cover?size=thumb","cover_alt":"Computational Intelligence Empowering Energy: A New Intelligent Paradigm from Data-Driven to Mechanism-Integrated","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4226","source_checked_at":"2026-10-30","agenda_json":"[[\"Data- and Physics-Driven Well Test Interpretation Software\",\"庞善臣\"],[\"Methods and Practice for Constructing High-Quality Energy Datasets\",\"宋亮\"],[\"From Data-Driven to World Models: A New Paradigm for AI-Empowered New Power Systems\",\"张莹\"],[\"Research Progress and Field Applications of Intelligent Fracturing for Unconventional Reservoirs\",\"曾凡辉\"]]","code":"CNCC2026-4226","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-415","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4226","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"计算智能赋能能源：从数据驱动到机理融合的智能化新范式","original_topic":"未分类","original_description":"随着人工智能与能源产业的深度融合，计算智能正成为驱动能源行业高质量发展的新质生产力。本论坛以“计算智能赋能能源：从数据驱动到机理融合的智能化新范式”为主题，聚焦能源智能化蓬勃发展期的前沿挑战与机遇。论坛将围绕能源多源异构数据治理与知识挖掘、复杂工业场景下的智能感知与认知、基于物理信息神经网络（PINNs）及大模型的机理与数据融合建模，以及生产决策优化等核心议题展开深入探讨。旨在搭建一个跨学科的交流平台，汇聚计算机科学与能源工程领域的专家学者，共同探索计算智能在能源全产业链中的创新应用路径，推动能源行业的数字化转型与智能化升级。","people":[{"forum_id":4226,"role":"chair","name":"彭博","affiliation":"西南石油大学","sort_order":0},{"forum_id":4226,"role":"guest","name":"庞善臣","affiliation":"中国石油大学（华东）","sort_order":0},{"forum_id":4226,"role":"guest","name":"宋亮","affiliation":"清华四川能源互联网研究院大数据研究所","sort_order":1},{"forum_id":4226,"role":"guest","name":"张莹","affiliation":"华北电力大学","sort_order":2},{"forum_id":4226,"role":"guest","name":"曾凡辉","affiliation":"西南石油大学","sort_order":3},{"forum_id":4226,"role":"guest","name":"李平","affiliation":"西南石油大学","sort_order":4}],"host":"","image_url":"/api/forums/4226/cover"},{"id":4227,"title":"From Interaction to Collaboration: Human-Computer Interaction in the Era of Embodied Intelligence","topic":"Robotics and Embodied Intelligence","kind":"专题论坛","chair":"佟馨","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"西博城5号馆-510","description":"With the accelerated integration of multimodal foundation models, agents, robots, and spatial computing, human-computer interaction is moving from information exchange on screens toward a new stage in which humans and agents jointly perceive, make collaborative decisions, and act together in real environments. Embodied intelligence not only means that agents have diverse physical carriers, but also emphasizes their ability to understand human intentions and collaboratively complete real-world tasks. Under this trend, humans, agents, and the environment together form a dynamically coupled collaborative interaction system. This forum will explore three key topics: through robots and multi-terminal intelligent devices, (1) how can embodied agents perceive humans and the environment more comprehensively? (2) how can embodied agents understand human intentions more accurately? (3) how can embodied agents better collaborate with humans and enhance human capabilities? The forum will bring together experts from industry, academia, and research to share the latest results and typical practices, and to jointly discuss the future evolution of human-machine relationships, interaction methods, and application ecosystems driven by embodied intelligence.","is_demo":0,"cover_url":"/api/forums/4227/cover?size=thumb","cover_alt":"From Interaction to Collaboration: Human-Computer Interaction in the Era of Embodied Intelligence","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4227","source_checked_at":"2026-10-30","agenda_json":"[[\"Some Thoughts on Developing Spatial Interactive System Software\",\"田丰\"],[\"The Last Millimeter of Embodied Intelligence: Key Technologies for Tactile-Feedback Teleoperation of Robotic Dexterous Hands\",\"王党校\"],[\"Embodied Intelligent Cockpit Interaction: Data-Driven, Smart Materials, and Collaborative Design\",\"米海鹏\"],[\"From Human-Like Memory to Embodied Intelligence: Exploring Personalized Cognition and Action\",\"张柳新\"],[\"From Manipulation to Collaboration: Productization Practices of Human-Computer Interaction in Commercial Service Scenarios in the Embodied Intelligence Era\",\"陈震\"]]","code":"CNCC2026-4227","forum_type":"专题论坛","format":null,"theme_code":"机器人与具身智能","theme_name":"机器人与具身智能","venue":"西博城5号馆","room":"西博城5号馆-510","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4227","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"从交互到协同：具身智能时代的人机交互","original_topic":"机器人与具身智能","original_description":"随着多模态大模型、智能体、机器人与空间计算加速融合，人机交互正从屏幕中的信息交换，迈向人与智能体在真实环境中共同感知、协同决策和联合行动的新阶段。具身智能不仅意味着智能体拥有多样化实体载体，更强调其能够理解人的意图并协同完成现实任务。在这一趋势下，人、智能体与环境共同构成动态耦合的协同交互系统。本论坛将围绕三个关键议题展开探讨：通过机器人和多终端智能设备（1）具身智能体如何更全面地感知人类与环境？（2）具身智能体如何更准确地理解人类意图？（3）具身智能体如何更好地与人协作并增强人的能力？论坛将汇聚产学研专家，分享最新成果与典型实践，共同探讨具身智能驱动下人机关系、交互方式与应用生态的未来演进。","people":[{"forum_id":4227,"role":"chair","name":"佟馨","affiliation":"香港科技大学（广州）","sort_order":0},{"forum_id":4227,"role":"guest","name":"田丰","affiliation":"中科院软件所","sort_order":0},{"forum_id":4227,"role":"guest","name":"王党校","affiliation":"北京航空航天大学","sort_order":1},{"forum_id":4227,"role":"guest","name":"米海鹏","affiliation":"清华大学","sort_order":2},{"forum_id":4227,"role":"guest","name":"张柳新","affiliation":"联想（北京）有限公司","sort_order":3},{"forum_id":4227,"role":"guest","name":"陈震","affiliation":"享刻智能技术（北京）有限公司","sort_order":4},{"forum_id":4227,"role":"guest","name":"范向民","affiliation":"北京中科睿医信息科技有限公司","sort_order":5},{"forum_id":4227,"role":"guest","name":"王重阳","affiliation":"四川大学","sort_order":6},{"forum_id":4227,"role":"guest","name":"张鑫磊","affiliation":"华为终端有限公司","sort_order":7},{"forum_id":4227,"role":"guest","name":"赵凯星","affiliation":"西北工业大学","sort_order":8},{"forum_id":4227,"role":"guest","name":"梁宸","affiliation":"香港科技大学（广州）","sort_order":9},{"forum_id":4227,"role":"guest","name":"龚江涛","affiliation":"上海交通大学","sort_order":10},{"forum_id":4227,"role":"guest","name":"卞玉龙","affiliation":"山东大学","sort_order":11}],"host":"","image_url":"/api/forums/4227/cover"},{"id":4241,"title":"Intelligent IoT in the Era of Large Models: Networks, Computing, and Trust","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"彭舰、张欢欢","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-406A兴隆湖A","description":"Large models are reshaping the paradigm of intelligent IoT. The shift from the past \"AI empowering IoT\" to today's \"IoT supporting AI\" is reconfiguring the system boundaries and capability division of labor for intelligent IoT in the era of large models. Cloud-edge-device collaboration, network-native intelligence, precise measurement, and trustworthy security are becoming the core pillars of next-generation systems. However, many open problems remain, including high-precision measurement of complex networks, cross-domain collaborative scheduling of sensing, transmission, and computing resources, and reliable transmission in high-throughput environments. This forum will focus on the key technologies of intelligent IoT in the era of large models, inviting authoritative scholars in the field to share their latest research results and to discuss in depth full-stack breakthroughs from algorithmic innovation to system deployment, driving the industry toward the era of \"intelligent connection of everything.\"","is_demo":0,"cover_url":"/api/forums/4241/cover?size=thumb","cover_alt":"Intelligent IoT in the Era of Large Models: Networks, Computing, and Trust","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4241","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Remarks\",\"彭舰\"],[\"Reflections on the Development of Computing Power Networks and Some Practical Explorations\",\"郭得科\"],[\"Full-Volume High-Frequency Network Visualization Technology\",\"谢鲲\"],[\"On-Device Intelligence: Overcoming Large Model Inference Challenges in Resource-Constrained Scenarios\",\"徐宏力\"],[\"Integration of Communication, Sensing, Computing, and Intelligence in 6G Mobile Communication Networks\",\"杨鲲\"],[\"Cross-Domain Distributed System Communication Optimization in the Large Model Era\",\"李彤\"]]","code":"CNCC2026-4241","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"天府国际会议中心","room":"天府国际会议中心-406A兴隆湖A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4241","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"大模型时代的智能物联：网络、计算与可信","original_topic":"新型计算与网络架构","original_description":"大模型正在重塑智能物联范式。从过去的“AI赋能物联”，演进为如今的“物联支撑AI”，这一变革正在重构大模型时代下智能物联的系统边界与能力分工。云边端协同、网络内生智能、精准测量与可信安全，正成为下一代系统的核心支柱。然而，复杂网络的高精度测量、感传算资源的跨域协同调度，以及高通量环境下的可靠传输，仍有很多未解之题。本论坛将聚焦大模型时代下智能物联的关键技术，邀请领域内权威学者分享最新研究成果，深入探讨从算法创新到系统落地的全栈突破，推动产业迈向“万物智联”时代。","people":[{"forum_id":4241,"role":"chair","name":"彭舰","affiliation":"四川大学","sort_order":0},{"forum_id":4241,"role":"guest","name":"郭得科","affiliation":"中山大学","sort_order":0},{"forum_id":4241,"role":"co_chair","name":"张欢欢","affiliation":"北京邮电大学","sort_order":1},{"forum_id":4241,"role":"guest","name":"谢鲲","affiliation":"湖南大学","sort_order":1},{"forum_id":4241,"role":"guest","name":"徐宏力","affiliation":"中国科学技术大学","sort_order":2},{"forum_id":4241,"role":"guest","name":"杨鲲","affiliation":"南京大学","sort_order":3},{"forum_id":4241,"role":"guest","name":"李彤","affiliation":"中国人民大学","sort_order":4},{"forum_id":4241,"role":"guest","name":"周安福","affiliation":"北京邮电大学计算机学院","sort_order":5},{"forum_id":4241,"role":"guest","name":"马华东","affiliation":"北京邮电大学","sort_order":6}],"host":"","image_url":"/api/forums/4241/cover"},{"id":4242,"title":"From AI Alchemists to Art Directors: AI-Driven Cross-Media Integration and Arts Talent Cultivation","topic":"General","kind":"专题论坛","chair":"白小墨、陆丽","starts_at":"2026-10-22T13:30:00+08:00","ends_at":"2026-10-22T17:30:00+08:00","place":"天府国际会议中心-106锦城","description":"In 2026, AI generation technology has crossed the boundaries of single media and shows a pronounced trend of \"modal fusion.\" The same Transformer architecture can drive the generation of text, images, audio, video and 3D models; the same set of diffusion models can generate a painting from a piece of text, then generate a soundtrack from that painting, and then generate a short video. Creators are no longer defined by identities such as \"I am a painter\" or \"I am a composer\"; instead, empowered by AI tools, they become cross-media \"art directors.\" However, current university arts education still centers on traditional disciplines (composition/performance/painting/sculpture/film and television), with curricula, evaluation and faculty all organized around a single medium, and no systematic cultivation paradigm for \"cross-media art directors\" has yet been formed. This forum will explore how to cultivate a new type of arts talent who both understands algorithms and is versed in multiple media?","is_demo":0,"cover_url":"/api/forums/4242/cover?size=thumb","cover_alt":"From AI Alchemists to Art Directors: AI-Driven Cross-Media Integration and Arts Talent Cultivation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4242","source_checked_at":"2026-10-30","agenda_json":"[[\"The Beauty of Image Generation\",\"刘家瑛\"],[\"AI-Empowered Cross-Media Game Creation: Cultivating Interdisciplinary Game Art Design Talent\",\"何俊\"],[\"Human-Machine Collaborative Art Perception Logic and Decision-Making Mechanisms in Creation\",\"王晓璇\"]]","code":"CNCC2026-4242","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"天府国际会议中心","room":"天府国际会议中心-106锦城","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":3,"forum_id":"CNCC2026-4242","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"从AI炼丹师到艺术主理人：AI驱动的跨媒介融合与艺术人才培养","original_topic":"综合","original_description":"2026年，AI生成技术已跨越单一媒介边界，呈现出显著的“模态融合”趋势。同一套Transformer架构可以驱动文本、图像、音频、视频、3D模型的生成；同一套扩散模型可以从一段文字生成一幅画、再从这幅画生成一段配乐、进而生成一支短视频。创作者不再被“我是画家”或“我是作曲家”的身份定义，而是在AI工具的赋能下成为跨媒介的“艺术主理人”。然而当前高校艺术教育仍以传统分科（作曲/表演/绘画/雕塑/影视）为核心，课程、评价、师资均按单一媒介组织，未形成面向\"跨媒介艺术主理人\"的系统培养范式。本论坛将探讨如何培养既懂算法又通多媒介的新型艺术人才？","people":[{"forum_id":4242,"role":"chair","name":"白小墨","affiliation":"成都美术学院","sort_order":0},{"forum_id":4242,"role":"guest","name":"苏彦聪","affiliation":"厦门理工学院","sort_order":0},{"forum_id":4242,"role":"co_chair","name":"陆丽","affiliation":"嵩山实验室","sort_order":1},{"forum_id":4242,"role":"guest","name":"党锐","affiliation":"成都工业学院","sort_order":1},{"forum_id":4242,"role":"guest","name":"刘家瑛","affiliation":"北京大学","sort_order":2},{"forum_id":4242,"role":"guest","name":"何俊","affiliation":"福州大学","sort_order":3},{"forum_id":4242,"role":"guest","name":"王晓璇","affiliation":"沈阳音乐学院","sort_order":4}],"host":"","image_url":"/api/forums/4242/cover"},{"id":4298,"title":"Plenary Forum (Oct 23) — The Future of the Software Industry in the Age of Intelligent Programming","topic":"Unclassified","kind":"大会论坛","chair":"黄罡（主持）","starts_at":"2026-10-23T11:10:00+08:00","ends_at":"2026-10-23T12:10:00+08:00","place":"西博城1号馆","description":"When large models and AI Agents can automatically generate high-quality code, \"intelligent programming\" has begun to become the core engine of transformation in the software industry. China attaches great importance to the integrated development of artificial intelligence and the software industry, treating it as an important engine for promoting the deep integration of the digital economy and the real economy and for seizing the high ground in the intelligent era. Against this background, the conference forum, with the theme \"The Future of the Software Industry in the Era of Intelligent Programming,\" specially invites academicians, scholars, and industry experts in fields such as software engineering, software education, artificial intelligence, and industrial practice to engage in in-depth dialogue around key topics such as intelligent programming reshaping the software industry, the opportunities and challenges of programming for everyone, and the integration of education, science and technology, and talent in the software field. The forum aims to build consensus and clarify paths, and to contribute wisdom and solutions for promoting the high-quality development of China's software industry and for the transformation and upgrading of all industries in the intelligent era. Everyone is welcome to pay attention.","is_demo":0,"cover_url":"/api/forums/4298/cover?size=thumb","cover_alt":"Plenary Forum (Oct 23) — The Future of the Software Industry in the Age of Intelligent Programming","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4298","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4298","forum_type":"大会论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城1号馆","room":"西博城1号馆","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[\"官网标注主持，未标注主席；主持与主席分开存储。\"]","agenda_count":0,"forum_id":"CNCC2026-4298","legacy_id":1,"lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"23日大会论坛——智能编程时代的软件产业未来","original_topic":"未分类","original_description":"当大模型与智能体能够自动生成高质量代码，“智能编程”开始成为软件产业变革的核心引擎。我国高度重视人工智能与软件产业的融合发展，将其作为推动数字经济和实体经济深度融合、抢占智能化时代制高点的重要引擎。在此背景下，大会论坛以“智能编程时代的软件产业未来”为主题，特别邀请软件工程、软件教育、人工智能、产业实践等领域的院士学者与业内专家，围绕智能编程重构软件产业、人人皆能编程的机遇与挑战、软件领域教育科技人才一体化等关键议题展开深度对话。论坛旨在凝聚共识、明晰路径，为推动我国软件产业高质量发展、为千行百业在智能化时代的转型升级贡献智慧与方案。欢迎大家共同关注。","people":[{"forum_id":4298,"role":"host","name":"黄罡","affiliation":"北京大学信息科学技术学院","sort_order":0},{"forum_id":4298,"role":"guest","name":"王怀民","affiliation":"国防科技大学计算机学院","sort_order":0},{"forum_id":4298,"role":"guest","name":"李戈","affiliation":"北京大学","sort_order":1},{"forum_id":4298,"role":"guest","name":"唐杰","affiliation":"清华大学","sort_order":2},{"forum_id":4298,"role":"guest","name":"安筱鹏","affiliation":"阿里云智能集团","sort_order":3}],"host":"黄罡","image_url":"/api/forums/4298/cover"},{"id":4002,"title":"Second Industrial Intelligence Forum: Theory and Systems of Industrial AI Agents","topic":"AI for Industry","kind":"专题论坛","chair":"王忠杰","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-115A九寨厅A","description":"The first Industrial Intelligence Forum was recognized as an \"Outstanding Special Forum of CNCC 2025.\" Building on that foundation, the Second Industrial Intelligence Forum at CNCC 2026 continues and upgrades the discussion, with the theme \"Theory and Systems of Industrial AI Agents.\" It focuses on how generative artificial intelligence is driving industrial intelligence from tool assistance toward goal-driven, autonomous action and multi-agent collaboration, connecting data, models, knowledge, software, equipment, and human engineers. Compared with general-purpose agents, industrial sites are characterized by multi-source heterogeneous data, fragmented knowledge, strongly constrained processes, and clear system safety boundaries. They face problems such as insufficient AI Ready Data, difficulty unifying industrial large models with domain knowledge, the lack of describable, composable, and deliverable mechanisms for multi-agent collaboration, and incomplete value validation, responsibility boundaries, and trusted governance systems. This forum upgrades the first forum's discussion of industrial large models, intelligent services, and industrial software into a complete chain of topics spanning \"theory-system-site,\" examining the foundational theory of industrial agents, industrial large models and digital genealogies, agent service computing, and industrial software integration, and, through highly constrained scenarios such as aviation manufacturing, equipment engineering, and source-grid-load-storage, exploring the path from laboratory prototypes to operable, auditable, and evolvable industrial systems.","is_demo":0,"cover_url":"/api/forums/4002/cover?size=thumb","cover_alt":"Second Industrial Intelligence Forum: Theory and Systems of Industrial AI Agents","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4002","source_checked_at":"2026-10-30","agenda_json":"[[\"Industrial Agents: Development Trends, Scientific Problems, and Frontier Explorations\",\"王忠杰\"],[\"Industrial Large Models + Agents + Digital Genealogy: Driving the Future Industrial World\",\"任磊\"],[\"Source-Grid-Load-Storage Collaborative Scheduling and Predictive Maintenance Technologies\",\"钟伟民\"],[\"Agent Service Computing: Research Framework and Industrial Scenario Practice\",\"邓水光\"],[\"Building an AI Ready Data System for Agents\",\"王晨\"],[\"Exploration and Application of System Computing Simulation Industrial Software and Equipment Engineering Agents\",\"周凡利\"]]","code":"CNCC2026-4002","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"天府国际会议中心","room":"天府国际会议中心-115A九寨厅A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4002","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第二届工业智能论坛：工业智能体理论与系统","original_topic":"AI+行业","original_description":"首届工业智能论坛获评“CNCC 2025优秀专题论坛”。CNCC 2026第二届工业智能论坛在此基础上延续并升级，以“工业智能体理论与系统”为主题，聚焦生成式人工智能推动工业智能从工具辅助迈向目标驱动、自主行动与多智能体协同，贯通数据、模型、知识、软件、设备与人类工程师。与通用智能体相比，工业现场具有数据多源异构、知识碎片化、流程强约束和系统安全边界明确等特征。面临AI Ready Data不足，工业大模型与领域知识难以统一，多智能体协同缺少可描述、可组合、可交付机制，以及价值验证、责任边界和可信治理体系不完善等问题。本届论坛将首届对工业大模型、智能服务与工业软件的讨论，升级为“理论-系统-现场”的完整议题链，研讨工业智能体基础理论、工业大模型与数字族谱、智能体服务计算和工业软件集成，并通过航空制造、装备工程、源网荷储等高约束场景，探索从实验室原型走向可运行、可审计、可演化工业系统的路径。","people":[{"forum_id":4002,"role":"chair","name":"王忠杰","affiliation":"哈尔滨工业大学","sort_order":0},{"forum_id":4002,"role":"guest","name":"任磊","affiliation":"北京航空航天大学","sort_order":0},{"forum_id":4002,"role":"guest","name":"钟伟民","affiliation":"华东理工大学","sort_order":1},{"forum_id":4002,"role":"guest","name":"邓水光","affiliation":"浙江大学","sort_order":2},{"forum_id":4002,"role":"guest","name":"王晨","affiliation":"清华大学","sort_order":3},{"forum_id":4002,"role":"guest","name":"周凡利","affiliation":"苏州同元软控技术股份有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4002/cover"},{"id":4004,"title":"The 8th CNCC NOI Forum — With AI as a Mirror, What Is the Essence and Future of Informatics Olympiad?","topic":"General","kind":"专题论坛","chair":"韩文弢、赵启阳","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城2号馆-208","description":"The National Olympiad in Informatics (NOI), as a benchmark activity for cultivating computational thinking among Chinese youth, has over more than forty years trained and selected a large number of outstanding reserve talents and has won widespread recognition and high praise. Currently, new artificial intelligence technologies (AI) represented by large language models (LLM) have acquired the ability to solve some algorithm competition problems. Against this background, the NOI community needs to think deeply about the essence of the Informatics Olympiad, attempt to clarify its future development direction, and respond to the widespread concerns of Informatics Olympiad teachers and students. Among these, the cultivation of computational thinking has become the top priority. With the theme \"The Essence and Future of the Informatics Olympiad: With AI as a Mirror,\" the forum will discuss the following questions:\n1. When machines can also compete in contests, where lies the core competitiveness of Informatics Olympiad participants?\n2. As AI tools become increasingly easy to use, how should the teaching, learning, and assessment of the Informatics Olympiad evolve?\n3. How can young people be guided to develop computational thinking?","is_demo":0,"cover_url":"/api/forums/4004/cover?size=thumb","cover_alt":"The 8th CNCC NOI Forum — With AI as a Mirror, What Is the Essence and Future of Informatics Olympiad?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4004","source_checked_at":"2026-10-30","agenda_json":"[[\"AI as a Mirror: The \\\"Constants\\\" and \\\"Adaptations\\\" of Informatics Olympiads\",\"金靖\"],[\"From Structural Insight to Generalization Ability: Rethinking the Educational Goals of Informatics Olympiads in the AI Era\",\"肖然\"],[\"AI as a Mirror, Returning to the Essence of Informatics Olympiads—Strengthening the Core of Thinking, Reshaping the Educational Model\",\"蔺洋\"],[\"Returning to Fundamentals: How High School Students Can Begin Their Journey in Artificial Intelligence\",\"郑子杰\"],[\"Paths and Implementation of Region-Oriented AI Innovation Talent Cultivation\",\"谭洪政\"],[\"Exploration of Computational Thinking Education for University Engineering Students\",\"韩文弢\"]]","code":"CNCC2026-4004","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城2号馆","room":"西博城2号馆-208","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4004","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第八届CNCC NOI论坛——以 AI 为镜，信息学奥赛的本质与未来是什么？","original_topic":"综合","original_description":"全国青少年信息学奥林匹克（NOI）作为我国青少年计算思维培养方面的标杆性活动，四十多年来培养和选拔了一大批优秀的后备人才，赢得了广泛认可和高度赞誉。当前以大语言模型（LLM）为代表的新型人工智能技术（AI），具备了解决部分算法竞赛题目的能力。在此背景下，NOI社区需要深入思考信息学奥赛的本质，尝试厘清未来的发展方向，回应信息学奥赛师生的广泛关切。其中，计算思维的培养成为重中之重。论坛将以“信息学奥赛的本质与未来：以AI为镜”为主题，围绕以下问题展开讨论：\n1、当机器也能打竞赛，信奥选手的核心竞争力在哪里？\n2、AI工具越来越好用，信奥教、学、考应该如何演进？\n3、如何引导青少年形成计算思维？","people":[{"forum_id":4004,"role":"chair","name":"韩文弢","affiliation":"清华大学","sort_order":0},{"forum_id":4004,"role":"guest","name":"金靖","affiliation":"华东师范大学第二附属中学","sort_order":0},{"forum_id":4004,"role":"co_chair","name":"赵启阳","affiliation":"北京航空航天大学","sort_order":1},{"forum_id":4004,"role":"guest","name":"肖然","affiliation":"北京大学附属中学","sort_order":1},{"forum_id":4004,"role":"guest","name":"蔺洋","affiliation":"成都市第七中学","sort_order":2},{"forum_id":4004,"role":"guest","name":"郑子杰","affiliation":"北京市十一学校","sort_order":3},{"forum_id":4004,"role":"guest","name":"谭洪政","affiliation":"清华大学附属中学","sort_order":4},{"forum_id":4004,"role":"guest","name":"谢秋锋","affiliation":"湖南省长沙市长郡中学","sort_order":5}],"host":"","image_url":"/api/forums/4004/cover"},{"id":4006,"title":"Mathematical and Theoretical Foundations of Large Models","topic":"Machine Learning Theory and Mechanisms","kind":"专题论坛","chair":"魏哲巍","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城2号馆-203","description":"Large models are constantly pushing the boundaries of capability, but why their capabilities emerge, why training is effective, when models are reliable, and where they fail remain fundamental questions that urgently need answers. The CNCC forum \"Mathematical and Theoretical Foundations of Large Models\" brings together experts in mathematics, statistics, and machine learning to focus on core topics such as representation and approximation, training and optimization, generalization and emergence, and interpretability and trustworthiness. Starting from frontier directions including symmetry, learning theory, compression perspectives, algebraic geometry, and failure boundaries, the forum explores the sources, mechanisms, and theoretical boundaries of large-model capabilities, providing new theoretical coordinates for building interpretable, trustworthy, and sustainable large models.","is_demo":0,"cover_url":"/api/forums/4006/cover?size=thumb","cover_alt":"Mathematical and Theoretical Foundations of Large Models","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4006","source_checked_at":"2026-10-30","agenda_json":"[[\"Symmetry: A Principle That AI Modeling Should Follow\",\"谢琦\"],[\"Machine Learning Theory in the Era of Large Models\",\"王立威\"],[\"Understanding Large Language Model Behavior Through a Compression Lens: Knowledge Acquisition, Scaling Laws, and In-Context Learning\",\"李建\"],[\"The Expressiveness Boundaries of Large Models from an Algebraic Geometry Perspective\",\"刘勇\"],[\"Beyond Risk Metrics: The Information Boundaries of AI Failures\",\"张俊杰\"]]","code":"CNCC2026-4006","forum_type":"专题论坛","format":null,"theme_code":"机器学习理论与机理","theme_name":"机器学习理论与机理","venue":"西博城2号馆","room":"西博城2号馆-203","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4006","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"大模型的数理基础","original_topic":"机器学习理论与机理","original_description":"大模型正在不断刷新能力边界，但其能力为何涌现、训练为何有效、模型何时可靠又会在何处失效，仍是亟待回答的基础问题。CNCC“⼤模型的数理基础”论坛汇聚数学、统计与机器学习领域专家，围绕表示与逼近、训练与优化、泛化与涌现、可解释性与可信性等核心议题，从对称性、学习理论、压缩视角、代数几何及失效边界等前沿方向切入，探寻大模型能力的来源、机制与理论边界，为构建可解释、可信赖、可持续的大模型提供新的理论坐标。","people":[{"forum_id":4006,"role":"chair","name":"魏哲巍","affiliation":"中国人民大学","sort_order":0},{"forum_id":4006,"role":"guest","name":"谢琦","affiliation":"西安交通大学","sort_order":0},{"forum_id":4006,"role":"guest","name":"王立威","affiliation":"北京大学","sort_order":1},{"forum_id":4006,"role":"guest","name":"李建","affiliation":"清华大学","sort_order":2},{"forum_id":4006,"role":"guest","name":"刘勇","affiliation":"中国人民大学","sort_order":3},{"forum_id":4006,"role":"guest","name":"张俊杰","affiliation":"西安交通大学","sort_order":4},{"forum_id":4006,"role":"guest","name":"孟德宇","affiliation":"西安交通大学","sort_order":5}],"host":"","image_url":"/api/forums/4006/cover"},{"id":4007,"title":"Frontier Exploration and Optimization of Intelligent Computing System Software","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"王新","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-109自流井","description":"From dialogue assistants to agent applications, from content creation to scientific research, large models have been deployed at scale and have become foundational capabilities empowering all industries. Behind all this, the continuous support of intelligent computing system software in computing power, storage, and communication is indispensable.\n\nHowever, the large-scale deployment of large models has brought many challenges to intelligent computing systems. On the training side, ultra-large-scale clusters have sharply amplified the pressure on communication, storage, and fault tolerance. On the inference side, memory access bandwidth and the utilization efficiency of heterogeneous computing power have become new bottlenecks. At the same time, the system software built to address these challenges has itself become increasingly complex, and its design, verification, and evolution can no longer be managed by traditional methods alone.\n\nCentering on the question \"what kind of intelligent computing system support does the large-scale deployment of large models require\", this forum invites experts from academia and industry to discuss: on the training side, checkpointing and communication optimization for large-model parallel training; on the inference side, fine-grained inference systems for heterogeneous hardware, speculative decoding and hardware-software co-design to break through memory bottlenecks, and storage-compute collaborative scheduling under serverless architectures; at the level of system complexity, the new paradigm of \"system intelligence\" that uses AI to reshape system design and verification, jointly promoting the continuous evolution of intelligent computing systems in the era of large models.","is_demo":0,"cover_url":"/api/forums/4007/cover?size=thumb","cover_alt":"Frontier Exploration and Optimization of Intelligent Computing System Software","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4007","source_checked_at":"2026-10-30","agenda_json":"[[\"System Intelligence: The Next Frontier of Computer Systems\",\"周礼栋\"],[\"System Optimization Techniques in Large Model Parallel Training\",\"李东升\"],[\"Fine-Grained Heterogeneous Large Model Inference Systems\",\"王晨曦\"],[\"Serverless Architecture-Based Compute-Storage Collaborative Scheduling for Large Model Inference\",\"陈游旻\"],[\"Large Model Inference Optimization for Breaking Through Memory Bottlenecks\",\"李萌\"]]","code":"CNCC2026-4007","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"天府国际会议中心","room":"天府国际会议中心-109自流井","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4007","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智算系统软件前沿探索及优化","original_topic":"新型计算与网络架构","original_description":"从对话助手到智能体应用，从内容创作到科学研究，大模型已被大规模部署，成为赋能千行百业的基础能力。在这一切的背后，离不开智算系统软件在算力、存储与通信上的持续支撑。\n\n然而，大模型的规模化部署给智算系统带来了诸多挑战。在训练侧，超大规模集群给通信、存储与容错带来的压力急剧放大。在推理侧，访存带宽与异构算力的利用效率成为新的瓶颈。与此同时，为应对这些挑战而构建的系统软件本身也日益复杂，其设计、验证与演进已难以仅靠传统方法驾驭。\n\n本论坛围绕“大模型的规模化部署需要什么样的智算系统支撑”这一问题，邀请学术界与产业界专家展开研讨：在训练侧，探讨大模型并行训练的检查点与通信优化；在推理侧，探讨面向异构硬件的细粒度推理系统、突破内存瓶颈的推测解码与软硬协同设计，及Serverless架构下的存算协同调度；在系统复杂性层面，探讨以AI重塑系统设计与验证的“系统智能”新范式，共同推动大模型时代智算系统的持续演进。","people":[{"forum_id":4007,"role":"chair","name":"王新","affiliation":"复旦大学","sort_order":0},{"forum_id":4007,"role":"guest","name":"周礼栋","affiliation":"微软亚洲研究院","sort_order":0},{"forum_id":4007,"role":"guest","name":"李东升","affiliation":"国防科技大学","sort_order":1},{"forum_id":4007,"role":"guest","name":"王晨曦","affiliation":"中国科学院计算技术研究所","sort_order":2},{"forum_id":4007,"role":"guest","name":"陈游旻","affiliation":"上海交通大学","sort_order":3},{"forum_id":4007,"role":"guest","name":"李萌","affiliation":"北京大学","sort_order":4},{"forum_id":4007,"role":"guest","name":"沈家诚","affiliation":"昆山杜克大学","sort_order":5},{"forum_id":4007,"role":"guest","name":"罗旭川","affiliation":"微软亚洲研究院","sort_order":6},{"forum_id":4007,"role":"guest","name":"周扬帆","affiliation":"复旦大学","sort_order":7}],"host":"","image_url":"/api/forums/4007/cover"},{"id":4010,"title":"Building an Integrated Large-Model AI Education System Across Primary, Secondary, and Higher Education Based on LMCC Certification Talent Evaluation","topic":"General","kind":"专题论坛","chair":"文继荣","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城2号馆-201","description":"The rapid iteration of large models is driving the upgrading of the AI industry, and market demand for talent has expanded beyond traditional algorithm research and development to diverse tracks such as model application, engineering implementation, software-hardware collaboration, agent development, model deployment, and industry transformation. Facing this structural change in industry talent, LMCC large-model capability certification provides standardized, tiered capability assessment standards for all educational stages, filling the gap left by the lack of a unified evaluation system in AI education.\nThis forum focuses on the contemporary proposition of \"how to systematically cultivate large-model talent in primary, secondary, and higher education,\" bringing together experts from all sides to directly address the challenge of reconstructing the new AI talent capability map, and to discuss in depth the integration of the LMCC certification system, an integrated curriculum system spanning primary, secondary, and higher education, experimental platforms, teacher standards, and industry-education integration paths. Forum topics: how to use LMCC foundational-level certification standards to plant the seeds of digital literacy for \"mastering models\" in primary and secondary schools; how to benchmark against the LMCC certification system to forge large-model [talent] in higher education and vocational education stages","is_demo":0,"cover_url":"/api/forums/4010/cover?size=thumb","cover_alt":"Building an Integrated Large-Model AI Education System Across Primary, Secondary, and Higher Education Based on LMCC Certification Talent Evaluation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4010","source_checked_at":"2026-10-30","agenda_json":"[[\"Capability Advancement and Integrated Teaching of Large-Model Agents\",\"邱锡鹏\"],[\"The Future of Agent Programming\",\"谭焜\"],[\"Large-Model-Based Social Simulation\",\"陈旭\"],[\"How Primary and Secondary School Students Can Embrace AI\",\"蒋炎岩\"],[\"The Impact of LMCC on Secondary School Information Technology Courses\",\"李曙\"]]","code":"CNCC2026-4010","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城2号馆","room":"西博城2号馆-201","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4010","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"依托LMCC认证人才评价，构建贯通大中小学大模型AI教育体系","original_topic":"综合","original_description":"大模型快速迭代推动AI产业升级，市场人才的需求已突破传统算法研发，延伸至模型应用、工程落地、软硬件协同、智能体开发、模型部署与行业转化等多元赛道。面对这一产业人才结构变革，LMCC大模型能力认证提供全学段分层标准化能力考核标准，补齐AI教育统一评价体系缺失的短板。\n本论坛聚焦“大模型人才在大中小学如何系统培养”这一时代命题，汇聚各方专家，直面全新AI人才能力图谱重构难题，深度探讨融合LMCC认证体系、贯通大中小学一体化的课程体系、实验平台、师资标准与产教融合路径。论坛议题：如何依托LMCC基础级认证标准在中小学播下“驾驭模型”的数字素养种子；如何对标LMCC认证体系，在高校与职教阶段锻造大模型","people":[{"forum_id":4010,"role":"chair","name":"文继荣","affiliation":"中国人民大学","sort_order":0},{"forum_id":4010,"role":"guest","name":"邱锡鹏","affiliation":"复旦大学","sort_order":0},{"forum_id":4010,"role":"guest","name":"谭焜","affiliation":"华为技术有限公司","sort_order":1},{"forum_id":4010,"role":"guest","name":"陈旭","affiliation":"中国人民大学","sort_order":2},{"forum_id":4010,"role":"guest","name":"蒋炎岩","affiliation":"南京大学","sort_order":3},{"forum_id":4010,"role":"guest","name":"李曙","affiliation":"南京外国语学校","sort_order":4}],"host":"","image_url":"/api/forums/4010/cover"},{"id":4018,"title":"The 2nd Workshop on Super-Heterogeneous Computer Architecture","topic":"Unclassified","kind":"专题论坛","chair":"李雪琦","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-405","description":"Facing complex AI computing scenarios, current computing systems are challenged by the deep interweaving of multiple workloads such as compute-intensive, memory-intensive, and real-time environment interaction. Traditional homogeneous/heterogeneous computing architectures are difficult to support the above complex scenarios end-to-end and with high efficiency. With \"super-heterogeneous\" as its core concept, the forum explores the collaborative design of multiple computing components such as CPU/XPU/DSP/PIM at the circuit, architecture, and system levels, promoting theoretical innovation and technological breakthroughs in super-heterogeneous computer architecture. In the discussions of the previous forum, three types of super-heterogeneity were preliminarily clarified: type (such as CPU+GPU+LPU), location (such as compute-in-memory), and mechanism (such as quantum). On this basis, this forum will focus on AI agent and physical AI scenarios, as well as compute-in-memory super-heterogeneity, with emphasis on AI-assisted circuit design of key components and hardware-software design in compute-in-memory super-heterogeneity and other fundamental frontier issues, promoting the formation of a reference design for physical AI super-heterogeneous acceleration architecture.","is_demo":0,"cover_url":"/api/forums/4018/cover?size=thumb","cover_alt":"The 2nd Workshop on Super-Heterogeneous Computer Architecture","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4018","source_checked_at":"2026-10-30","agenda_json":"[[\"CPU Microarchitecture Design Enhanced by Code Hints Generated via Large Models\",\"徐策羽\"],[\"Enlightenment: Fully Automated Processor Logic Design\",\"承书尧\"],[\"Software-Hardware Co-Designed Heterogeneous Compute-in-Memory Fusion Architecture\",\"陈家贤\"],[\"Exploration of Compute-in-Memory Fusion Architecture Design for Domain-Fusion Intelligent Computing Workloads\",\"黄禹\"],[\"A Preliminary Exploration of the Compute-in-Memory Physical AI Processor (Panda)\",\"李雪琦\"]]","code":"CNCC2026-4018","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-405","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4018","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第二届超异构计算机体系结构研讨会","original_topic":"未分类","original_description":"面向复杂AI计算场景，当前计算系统面临计算密集、访存密集、实时环境交互等多种负载深度交织的挑战。传统的同构/异构计算架构难以端到端、高效能支撑上述复杂场景。论坛以\"超异构\"为核心理念，探索CPU/XPU/DSP/PIM等多种计算部件在电路、架构、系统等层面的协同设计，推动超异构计算机体系结构的理论创新与技术突破。上届论坛研讨中，初步厘清了类型（如CPU+GPU+LPU）、位置（如存算）、机理（如量子）三种超异构类型。在此基础上，本届论坛将聚焦智能体与物理AI场景，以及存算一体超异构，重点研讨AI辅助关键部件电路设计，存算超异构中的软硬件设计等基础前沿问题，推动形成物理AI超异构加速结构参考设计。","people":[{"forum_id":4018,"role":"chair","name":"李雪琦","affiliation":"中国科学院计算技术研究所","sort_order":0},{"forum_id":4018,"role":"guest","name":"徐策羽","affiliation":"香港科技大学","sort_order":0},{"forum_id":4018,"role":"guest","name":"承书尧","affiliation":"中国科学院计算技术研究所","sort_order":1},{"forum_id":4018,"role":"guest","name":"陈家贤","affiliation":"北京理工大学（珠海）","sort_order":2},{"forum_id":4018,"role":"guest","name":"黄禹","affiliation":"华中科技大学","sort_order":3},{"forum_id":4018,"role":"guest","name":"包云岗","affiliation":"中国科学院计算技术研究所","sort_order":4},{"forum_id":4018,"role":"guest","name":"武延军","affiliation":"中国科学院软件研究所","sort_order":5},{"forum_id":4018,"role":"guest","name":"李东升","affiliation":"国防科技大学","sort_order":6},{"forum_id":4018,"role":"guest","name":"金海","affiliation":"华中科技大学","sort_order":7},{"forum_id":4018,"role":"guest","name":"杜夏威","affiliation":"曙光信息产业（北京）有限公司","sort_order":8}],"host":"","image_url":"/api/forums/4018/cover"},{"id":4023,"title":"Second Forum on Adaptive Machine Learning in Non-stationary Environments: Challenges Brought by Large Models","topic":"Machine Learning Theory and Mechanisms","kind":"专题论坛","chair":"文益民","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-416","description":"The rise of large models has brought new challenges and opportunities to adaptive machine learning in non-stationary environments. First, large models are costly to train, and in non-stationary environments they cannot be frequently fully retrained, so their fixed knowledge struggles to respond promptly to changes in data distribution. Second, large models have enormous parameter scales, and when performing incremental learning on non-stationary data streams, they are highly prone to \"catastrophic forgetting.\" However, the powerful representation learning capability of large models can extract more generalizable invariant features from high-dimensional streaming data. With their outstanding in-context learning capability, large models can quickly adapt to new environments through carefully designed prompts or a small number of new samples. This forum will report and discuss the new paths, new problems and new insights that large models provide for adaptive machine learning in non-stationary environments.","is_demo":0,"cover_url":"/api/forums/4023/cover?size=thumb","cover_alt":"Second Forum on Adaptive Machine Learning in Non-stationary Environments: Challenges Brought by Large Models","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4023","source_checked_at":"2026-10-30","agenda_json":"[[\"Theory of Large Model Architecture Generalization and Hallucination Suppression\",\"侯臣平\"],[\"A Unified Perspective on Efficient Fine-Tuning of Large Models: Theoretical Analysis and Algorithm Design\",\"李晶晶\"],[\"Robust Data Stream Classification for Concept Drift: Research Progress and Open Problems in the Era of Large Models\",\"李培培\"],[\"Continuously Self-Evolving Agent Systems\",\"朱飞\"],[\"Uncertainty Analysis for Large Model Inference and Interaction\",\"张绍群\"]]","code":"CNCC2026-4023","forum_type":"专题论坛","format":null,"theme_code":"机器学习理论与机理","theme_name":"机器学习理论与机理","venue":"西博城4号馆","room":"西博城4号馆-416","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4023","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"非平稳环境下的自适应机器学习第二届论坛——大模型带来的挑战","original_topic":"机器学习理论与机理","original_description":"大模型的兴起为非平稳环境下的自适应机器学习带来了新的挑战与机遇。其一，大模型训练成本高昂，面对非平稳环境，无法频繁全量重训，其固化知识难以及时响应数据分布的变化。其二，大模型参数规模巨大，在非平稳数据流上进行增量学习时，极易引发“灾难性遗忘”。然而，大模型强大的表示学习能力，能从高维流数据中提取更具泛化性的不变特征。借助其突出的上下文学习能力，大模型可通过精心设计的提示或少量新样本快速适应新环境。本论坛将报告和讨论大模型给非平稳环境下的自适应机器学习提供的新路径，提出的新问题，带来的新启示。","people":[{"forum_id":4023,"role":"chair","name":"文益民","affiliation":"五邑大学","sort_order":0},{"forum_id":4023,"role":"guest","name":"侯臣平","affiliation":"国防科技大学","sort_order":0},{"forum_id":4023,"role":"guest","name":"李晶晶","affiliation":"电子科技大学","sort_order":1},{"forum_id":4023,"role":"guest","name":"李培培","affiliation":"合肥工业大学","sort_order":2},{"forum_id":4023,"role":"guest","name":"朱飞","affiliation":"中国科学院香港创新研究院","sort_order":3},{"forum_id":4023,"role":"guest","name":"张绍群","affiliation":"南京大学","sort_order":4}],"host":"","image_url":"/api/forums/4023/cover"},{"id":4027,"title":"Industrial AI for High-End Equipment Operations and Maintenance: Can Models Understand Physical Laws?","topic":"AI for Industry","kind":"专题论坛","chair":"魏宪、付昱凯","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-118B峨眉厅B","description":"High-end equipment such as high-speed trains, aircraft, tunnel boring machines, and water turbines forms the \"backbone\" of major national projects. These systems are highly coupled and have long faced the challenges of extreme service environments and zero-tolerance failure constraints. Traditional operations and maintenance (O&M) models are encountering systemic bottlenecks: for example, cross-scale physical degradation mechanisms are difficult to quantify, fragmented data can hardly characterize dynamic evolution, manual inspection is limited by experience gaps, and automated O&M based on fixed rules is trapped in \"cognitive blind spots\" regarding unknown failure modes.\n\tCurrently, industrial artificial intelligence represented by large models and agents is driving a transformation of the O&M paradigm toward \"dual-driven by data and knowledge.\" However, however, from the \"statistical impressiveness\" of the laboratory to \"high-reliability operation\" in industrial sites, the technological gap remains huge. Industrial sites face underlying challenges such as extreme scarcity of negative samples, impeded dynamic generalization, and difficulty reconciling data-driven approaches with physical consistency. The breakthrough of intelligent O&M for high-end equipment urgently requires moving beyond the fragmented application of single-point AI tools to build a full-chain, system-level solution of \"all-time perception-precise diagnosis-time-varying prediction-autonomous decision-closed-loop execution.\"\n\tOn this basis, this forum closely addresses the technological frontier and the pain points of industrial deployment, conducting in-depth deliberation around \"real pain points, real boundaries, and breakthrough paths\":\n\tTopic 1: Distinguishing real from pseudo pain points—where exactly does the \"pain\" of high-end equipment O&M lie? Is it that physical mechanisms are \"unclear,\" or that fragmented data are \"unreadable\"?\n\tAs the starting point of discussion, clarify the underlying pathologies of O&M. Is the predicament of traditional O&M because the cross-scale degradation laws of complex systems exceed the limits of human mechanistic cognition, or because multi-source heterogeneous data cannot effectively characterize existing laws? Does the introduction of agents achieve a paradigm reconstruction of O&M cognition, or merely replace it with a more advanced single-point tool patch?\n\tTopic 2: Distinguishing mechanism mining—under the \"physical islands\" of sample scarcity, do agents truly \"understand\" physical laws, or have they fallen into \"statistical illusions\"?\n\tStrike at the essence of the science. In extreme industrial scenarios where fault samples are extremely scarce and negative samples cannot be exhausted, are the emergent capabilities demonstrated by large models and agents based on a \"deep understanding\" of physical degradation mechanisms and process boundaries, or are they merely high-dimensional probabilistic fitting in a big-data space? Under zero-tolerance safety constraints, will industrial sites dare to trust \"black-box predictions\" that lack physical interpretability?\n\tTopic 3: Distinguishing engineering paradigms—from \"laboratory impressiveness\" to \"engineering reliability,\" should the breakthrough path lead to \"fully autonomous intelligence,\" rely on \"embedding physical mechanisms,\" or reconstruct \"human-machine collaboration\"?\n\tFacing future deployment, where exactly lies the path across the gap from accuracy to reliability? Should we adhere to purely data-driven approaches, hoping that agents will evolve into \"fully autonomous intelligence\" surpassing human experience; take the physical-consistency route, using mathematical means to hard-constrain physical equations into the model's底层 through \"mechanism embedding\"; or step outside the pure-algorithm perspective, position agents as \"copilots,\" and move toward reconstructing trust and safety defense mechanisms for \"human-machine collaboration\"? Among these three paradigms, which is the ultimate solution for industrial deployment?","is_demo":0,"cover_url":"/api/forums/4027/cover?size=thumb","cover_alt":"Industrial AI for High-End Equipment Operations and Maintenance: Can Models Understand Physical Laws?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4027","source_checked_at":"2026-10-30","agenda_json":"[[\"How Models \\\"Understand\\\" Physical Laws: Trustworthy Operation and Maintenance of High-End Equipment Driven by Physical Priors and Causal Inference\",\"程宏\"],[\"Noise-Robust and Uncertainty-Aware Intelligent Fault Diagnosis for Rotating Machinery: Letting Models \\\"Know What They Don't Know\\\"\",\"张晓光\"],[\"Research on Constructing Multimodal Large Models for Civil Aircraft Fault Diagnosis and Prediction\",\"王艺霖\"],[\"Green Intelligent Laser Manufacturing and Remanufacturing Technologies for High-Speed Transport Equipment\",\"张振林\"],[\"Application Research and Prospects of Intelligent Inspection Robots in Rail Transport Equipment\",\"李奕潇\"]]","code":"CNCC2026-4027","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"天府国际会议中心","room":"天府国际会议中心-118B峨眉厅B","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4027","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"工业人工智能落地高端装备运维：模型能读懂物理规律吗？","original_topic":"AI+行业","original_description":"高铁、飞机、盾构机、水轮机等高端装备，是国家重大工程的“硬脊梁”。其系统高度耦合，长期面临极端服役环境与零容忍失效约束的挑战。传统运维模式正遭遇系统性瓶颈：如跨尺度物理退化机理难以量化，碎片化数据难以表征动态演化，人工巡检受限于经验断层，而固定规则的自动化运维则困于未知失效模式的“认知盲区”。\n\t当前，以大模型和智能体为代表的工业人工智能正驱动运维范式向“数据与知识双驱动”变革。然而，然而，从实验室的“统计学惊艳”走向工业现场的“高可靠运行”，技术鸿沟依然巨大。工业现场面临负样本极度稀缺、动态泛化受阻，以及数据驱动与物理一致性难以调和等底层挑战。高端装备智能运维的突围，迫切需要跨越单点AI工具的碎片化应用，构建“全时感知-精准诊断-时变预测-自主决策-闭环执行”的全链条系统级解决方案。\n\t基于此，本论坛紧扣技术前沿与工业落地痛点，围绕“真痛点、真边界、破局路”展开深度思辨：\n\t议题一：痛点伪真之辨——高端装备运维到底“痛”在哪里？究竟是物理机理“说不清”，还是碎片数据“读不懂”？\n\t作为讨论起点，厘清运维的底层病灶。传统运维的困局，到底是复杂系统跨尺度退化规律超出了人类的机理认知极限，还是多源异构数据无法有效表征既有规律？智能体的引入，究竟是实现了运维认知的范式重构，还是仅仅换了个更高级的单点工具补丁？\n\t议题二：机理挖掘之辨——样本稀缺的“物理孤岛”下，智能体是真正“读懂”了物理规律，还是陷入了“统计学幻觉”？\n\t直击科学本质。在故障样本极度匮乏、负样本无法穷尽的极端工业场景下，大模型和智能体展现出的涌现能力，究竟是基于对物理退化机理与工艺边界的“深刻理解”，还是仅在大数据空间里进行的高维概率拟合？在零容忍的安全约束下，缺乏物理可解释性的“黑盒预测”，工业现场究竟敢不敢信？\n\t议题三：工程范式之辨——从“实验室惊艳”走向“工程可靠”，破局路径是走向“全自主智能”、依靠“物理机理嵌入”，还是重构“人机协同”？\n\t面向未来落地，跨越从精度到可靠性的鸿沟，究竟路在何方？是坚持纯数据驱动，寄希望于智能体演进为超越人类经验的“全自主智能”；是走物理一致性路线，通过数学手段将物理方程硬性约束至模型底层的“机理嵌入”；还是跳出纯算法视角，将智能体定位于“副驾驶”，走向“人机协同”的信任与安全防御机制重构？这三种范式谁才是工业落地的终极解？","people":[{"forum_id":4027,"role":"chair","name":"魏宪","affiliation":"华东师范大学","sort_order":0},{"forum_id":4027,"role":"guest","name":"程宏","affiliation":"上海立信会计金融学院","sort_order":0},{"forum_id":4027,"role":"co_chair","name":"付昱凯","affiliation":"中国科学院工业人工智能研究所","sort_order":1},{"forum_id":4027,"role":"guest","name":"张晓光","affiliation":"福耀科技大学","sort_order":1},{"forum_id":4027,"role":"guest","name":"王艺霖","affiliation":"上海浦江实验室","sort_order":2},{"forum_id":4027,"role":"guest","name":"张振林","affiliation":"西南交通大学","sort_order":3},{"forum_id":4027,"role":"guest","name":"李奕潇","affiliation":"南京景曜智能科技有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4027/cover"},{"id":4031,"title":"Agentic NET — Next-Generation Interconnect Architecture for AI Clusters: Opportunities and Challenges","topic":"Unclassified","kind":"专题论坛","chair":"王晓亮、管紫轩","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-401","description":"Artificial intelligence is undergoing a major paradigm shift. The Agent wave driven by OpenClaw is pushing AI from single-model interaction toward multi-agent autonomous collaboration, and the integrated network-computing-storage AI cluster is the core infrastructure supporting this transformation. As computing clusters expand toward a scale of more than one hundred thousand cards, building large-scale, efficient, and reliable intelligent computing networks adapted to the multi-agent era has become a key industry proposition.\nThis Agentic NET forum invites experts from Microsoft, Alibaba Cloud, NVIDIA, Tencent Cloud, and StepFun to interpret, in combination with practical deployments, the business characteristics of AI clusters, next-generation cluster interconnect architectures, communication optimization, and protocol innovation, and to analyze existing technical challenges and solutions; at the same time, it will discuss the development trends of ten-thousand-card cluster networks and open network design, explore breakthrough paths for key difficulties in GPU interconnection, and share cutting-edge industry perspectives, providing frontline practical references for researchers and engineering practitioners.","is_demo":0,"cover_url":"/api/forums/4031/cover?size=thumb","cover_alt":"Agentic NET — Next-Generation Interconnect Architecture for AI Clusters: Opportunities and Challenges","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4031","source_checked_at":"2026-10-30","agenda_json":"[[\"Practice and Reflections on the R&D and Design of Cubesandbox\",\"金峰\"],[\"Challenges and Practices of Cloud Networks in Agentic AI Applications\",\"宗志刚\"],[\"Challenges and Optimizations in Long-Context Training of Large Language Models\",\"Yuanwei Lu\"],[\"The Systematic Evolution of Baidu AI Networking\",\"李虎\"],[\"Efficient Next-Generation AI Network Infrastructure: From Scalable Network Architecture to Fast Simulation Platforms\",\"牛治雄\"]]","code":"CNCC2026-4031","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-401","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4031","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"Agentic NET- AI集群下一代互联架构、机遇与挑战","original_topic":"未分类","original_description":"人工智能迎来范式重大变革，OpenClaw 驱动的 Agent 浪潮推动 AI 从单模型交互走向多智能体自主协作，网算存一体化 AI 集群是支撑这场变革的核心基础设施。随着算力集群向超十万卡规模拓展，构建适配多智能体时代、大规模高效可靠的智算网络成为行业关键命题。\n本次 Agentic NET 论坛邀请微软、阿里云、英伟达、腾讯云、阶跃星辰专家，结合落地实践解读 AI 集群业务特征、下一代集群互联架构、通信优化与协议创新，剖析现存技术挑战与解决方案；同时探讨万卡集群网络发展趋势、开放网络设计，探索 GPU 互联关键难题破局路径，分享前沿产业观点，为科研人员与工程从业者提供一线实践参考。","people":[{"forum_id":4031,"role":"chair","name":"王晓亮","affiliation":"南京大学","sort_order":0},{"forum_id":4031,"role":"guest","name":"金峰","affiliation":"腾讯科技（深圳）有限公司","sort_order":0},{"forum_id":4031,"role":"co_chair","name":"管紫轩","affiliation":"华为","sort_order":1},{"forum_id":4031,"role":"guest","name":"宗志刚","affiliation":"阿里巴巴","sort_order":1},{"forum_id":4031,"role":"guest","name":"Yuanwei Lu","affiliation":"上海阶跃星辰智能科技有限公司","sort_order":2},{"forum_id":4031,"role":"guest","name":"李虎","affiliation":"百度时代网络技术（北京）有限公司","sort_order":3},{"forum_id":4031,"role":"guest","name":"牛治雄","affiliation":"微软亚洲研究院","sort_order":4}],"host":"","image_url":"/api/forums/4031/cover"},{"id":4048,"title":"Sixth Network Ecology Governance Forum: When AI Learns to \"Attack Autonomously\"—New Crises Brought by Agents and Mythos","topic":"AI Safety and Governance","kind":"专题论坛","chair":"虎嵩林、杨耀东","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-501","description":"The AI security field in 2026 has encountered a dual shock: autonomous Agents represented by OpenClaw are accelerating in application, and while empowering development, they also bring new risk surfaces such as unauthorized operations, Skill contamination, credential leakage and autonomous loss of control, making \"attack entry points everywhere\"; at the same time, a new generation of models represented by Mythos has demonstrated astonishing cybersecurity capabilities, and with the support of various Harnesses continues to refresh the highest level of automated vulnerability discovery and exploitation, with \"attack capabilities changing with each passing day.\" The complex superposition of the two is affecting the security paradigm of the AI era from the ground up and deeply reshaping the traditional security landscape.\nThis forum continues the previous editions' exploration of AI security issues, further focusing on the new security challenges in the era of \"Agents + Mythos,\" and invites experts from academia and industry to deliberate around three questions: What are the harms and causes of the dual shock and its superposition effects? What disruptive impact does the traditional security paradigm face; and how should a new security research and application landscape be constructed?","is_demo":0,"cover_url":"/api/forums/4048/cover?size=thumb","cover_alt":"Sixth Network Ecology Governance Forum: When AI Learns to \"Attack Autonomously\"—New Crises Brought by Agents and Mythos","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4048","source_checked_at":"2026-10-30","agenda_json":"[[\"On-Device Intelligence: Toward Safe and Controllable General Intelligence\",\"刘知远\"],[\"Analyzing Agent Security from the Data Dimension\",\"胡侠\"],[\"Evaluation and Evolution of Autonomous Attack-Defense Capabilities of AI Systems\",\"杨珉\"],[\"Small Models, Big Impact: From Specialized Models to a Verifiable Attack-Defense Loop\",\"黄龙涛\"],[\"From Thinking to Acting: Trustworthy AI in the Agent Era\",\"董胤蓬\"]]","code":"CNCC2026-4048","forum_type":"专题论坛","format":null,"theme_code":"AI安全与治理","theme_name":"AI安全与治理","venue":"西博城5号馆","room":"西博城5号馆-501","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4048","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第六届网络生态治理论坛：当AI学会“自主攻击”— Agent与Mythos带来的新危机","original_topic":"AI安全与治理","original_description":"2026年的人工智安全领域迎来了双重冲击：以OpenClaw为代表的自主Agent加速应用，在赋能发展的同时，也带来了越权操作、Skill污染、凭证泄露、自主失控等新型的风险面，使得“攻击入口无处不在”；与此同时，以Mythos为代表的新一代模型又展现出了令人惊叹的网络安全能力，在各类Harness的加持下持续刷新着自动化漏洞挖掘和利用的最高水平， “攻击能力日新月异”。二者的复杂叠加，正在从底层影响人工智能时代的安全范式，并深度地重塑着传统的安全格局。\n本届论坛延续前几届对人工智能安全议题的求索路线，进一步聚焦“智能体+Mythos“时代的新型安全挑战，邀请学界与产业界专家，围绕三大追问展开思辨：双重冲击及其叠加效应的危害与成因？传统的安全范式面临怎样的破坏性影响；新的安全研究与应用格局如何构建？","people":[{"forum_id":4048,"role":"chair","name":"虎嵩林","affiliation":"中国科学院信息工程研究所","sort_order":0},{"forum_id":4048,"role":"guest","name":"刘知远","affiliation":"清华大学","sort_order":0},{"forum_id":4048,"role":"guest","name":"胡侠","affiliation":"上海人工智能实验室","sort_order":1},{"forum_id":4048,"role":"co_chair","name":"杨耀东","affiliation":"北京大学","sort_order":1},{"forum_id":4048,"role":"guest","name":"杨珉","affiliation":"复旦大学计算与智能创新学院","sort_order":2},{"forum_id":4048,"role":"guest","name":"黄龙涛","affiliation":"阿里巴巴（北京）软件服务有限公司","sort_order":3},{"forum_id":4048,"role":"guest","name":"董胤蓬","affiliation":"清华大学","sort_order":4}],"host":"","image_url":"/api/forums/4048/cover"},{"id":4056,"title":"Satellite Computing: Key Technologies for Space-Based Computing Power","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"王尚广","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-403","description":"With the rapid iteration of artificial intelligence and the continuous rise in pressure on terrestrial computing power, energy, and network infrastructure, computing resources are gradually extending from the ground into space, and space-based computing power is thus becoming a new type of infrastructure for humanity to expand the boundaries of productivity. By deploying computing nodes on satellite platforms and relying on constellation-based networking, the computing, storage, and communication resources scattered in orbit are organized into a space-based computing system that can collaborate, be scheduled, and scale; satellites will no longer be merely information acquisition and transmission nodes, but are expected to further become space infrastructure that continuously provides intelligent processing and computing power services. This forum will focus on the core technologies and practical paths of space-based computing power infrastructure, inviting guests from related fields to give keynote speeches, with emphasis on typical directions such as autonomous constellation operation, intelligent resource scheduling, and converged network control, analyzing how satellites can enhance autonomous perception, autonomous decision-making, and anomaly handling capabilities in complex space environments, how to achieve coordinated scheduling of computing, communication, storage, and energy resources, and how to guarantee the stable operation of computing power services through inter-satellite, satellite-ground, and space-ground converged networks. Subsequently, the forum will hold an open discussion on the core topic of \"how to build an autonomous, efficient, and reliable space-based computing power infrastructure,\" sorting out the technological evolution path of satellite computing from ground dependence to in-orbit autonomy, from single-satellite processing to constellation collaboration, and from scattered resources to systematic computing power services, inspiring cross-disciplinary thinking among researchers in the computer and aerospace fields, and jointly promoting research and development in the field of space-based computing power.","is_demo":0,"cover_url":"/api/forums/4056/cover?size=thumb","cover_alt":"Satellite Computing: Key Technologies for Space-Based Computing Power","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4056","source_checked_at":"2026-10-30","agenda_json":"[[\"None\",\"王尚广\"],[\"Software-Defined Space-Air-Ground Integrated Network Control Mechanisms\",\"王兴伟\"],[\"Low-Overhead Measurement in Mega-Constellation Networks: Reflections from Terrestrial Networks to Space-Based Scenarios\",\"谢鲲\"],[\"Research on Collaborative Learning and Split Inference Between Satellite and Ground Large and Small Models\",\"高跃\"],[\"Research and Practice on Intelligent Autonomous Operation of Constellations\",\"李晴\"],[\"LLM-Agent Empowering Autonomous Satellite Computing Operations\",\"邢若粼\"]]","code":"CNCC2026-4056","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-403","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4056","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"卫星计算: 面向太空算力的关键技术","original_topic":"新型计算与网络架构","original_description":"随着人工智能快速迭代以及地面算力、能源和网络基础设施压力持续上升，计算资源正逐步从地面延伸至太空，太空算力由此成为人类拓展生产力边界的新型基础设施。通过将计算节点部署于卫星平台，并依托星座化组网，将分散在轨的计算、存储和通信资源组织为可协同、可调度、可扩展的天基计算系统，卫星将不再只是信息获取与传输节点，而有望进一步成为持续提供智能处理与算力服务的空间基础设施。本次论坛将围绕太空算力基础设施的核心技术与实践路径展开，邀请相关领域嘉宾作引导发言，重点探讨星座自主运行、智能资源调度和融合网络控制等典型方向，分析卫星如何在复杂空间环境中提升自主感知、自主决策与异常处置能力，如何实现计算、通信、存储和能源资源的协同调度，以及如何通过星间、星地和天地融合网络保障算力服务的稳定运行。随后，论坛将聚焦“如何构建自主、高效、可靠的太空算力基础设施”这一核心议题开展开放讨论，梳理卫星计算由地面依赖走向在轨自治、由单星处理走向星座协同、由资源分散走向系统化算力服务的技术演进路径，启发计算机与航天领域研究者的交叉思考，共同推动太空算力领域的研究与发展。","people":[{"forum_id":4056,"role":"chair","name":"王尚广","affiliation":"北京邮电大学","sort_order":0},{"forum_id":4056,"role":"guest","name":"王兴伟","affiliation":"东北大学","sort_order":0},{"forum_id":4056,"role":"guest","name":"谢鲲","affiliation":"湖南大学","sort_order":1},{"forum_id":4056,"role":"guest","name":"高跃","affiliation":"复旦大学","sort_order":2},{"forum_id":4056,"role":"guest","name":"李晴","affiliation":"北京邮电大学","sort_order":3},{"forum_id":4056,"role":"guest","name":"邢若粼","affiliation":"一苇宇航（北京）科技有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4056/cover"},{"id":4058,"title":"Autonomous but Bounded: Adversarial Attack-Defense and Security Governance for Agent Systems","topic":"AI Safety and Governance","kind":"专题论坛","chair":"曹亚男","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-408鹿溪河","description":"As large-model agents move from \"understanding and generation\" to \"planning and action,\" their autonomous capabilities in task decomposition, tool invocation, long-term memory, and multi-agent collaboration continue to strengthen, bringing security challenges that span the entire process of models, systems, and applications. With the theme \"Autonomous but Bounded: Adversarial Attack-Defense and Security Governance for Agent Systems,\" this forum focuses on how to make agent systems \"autonomous without losing control, acting without overstepping boundaries.\" The forum has invited renowned experts and scholars from domestic and international universities, research institutes, and industry to discuss risks such as jailbreak attacks on agent systems, memory poisoning, tool hijacking, privilege abuse, and collaborative loss of control, and to explore in depth key issues including tool and memory security, multi-agent collaboration protection, privacy and permission protection, security evaluation, and risk governance. Combined with specific application scenarios, the forum aims to promote the construction of an agent security assurance system that connects models, systems, and applications.","is_demo":0,"cover_url":"/api/forums/4058/cover?size=thumb","cover_alt":"Autonomous but Bounded: Adversarial Attack-Defense and Security Governance for Agent Systems","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4058","source_checked_at":"2026-10-30","agenda_json":"[[\"Security Risks and Challenges of Multi-Agent Systems\",\"曹亚男\"],[\"Risk Attribution and Protection of Agent Systems\",\"刘东瑞\"],[\"Security Attack and Defense and Trustworthy Application Challenges of Multimodal Large Models and Agents\",\"蔺琛皓\"],[\"From Large Models to Autonomous Agents: Blockchain-Enabled Verifiable Intelligent Collaboration Networks\",\"朱天清\"],[\"Security Practices and Exploration of Trustworthy Large-Model Agents\",\"王维强\"]]","code":"CNCC2026-4058","forum_type":"专题论坛","format":null,"theme_code":"AI安全与治理","theme_name":"AI安全与治理","venue":"天府国际会议中心","room":"天府国际会议中心-408鹿溪河","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4058","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"自主而有界：智能体系统的对抗攻防与安全治理","original_topic":"AI安全与治理","original_description":"随着大模型智能体从“理解与生成”迈向“规划与行动”，其在任务拆解、工具调用、长期记忆和多智能体协作中的自主能力不断增强，也带来了贯穿模型、系统与应用全流程的安全挑战。本论坛以“自主而有界：智能体系统的对抗攻防与安全治理”为主题，聚焦如何让智能体系统“自主而不失控、行动而不越界”。论坛邀请了来自国内外高校、科研院所和产业界的知名专家学者，围绕智能体系统越狱攻击、记忆投毒、工具劫持、权限滥用和协同失控等风险展开研讨，深入探讨工具与记忆安全、多智能体协作防护、隐私权限保护、安全评测与风险治理等关键问题，并结合具体应用场景，推动构建贯通模型、系统与应用的智能体安全保障体系。","people":[{"forum_id":4058,"role":"guest","name":"刘东瑞","affiliation":"上海人工智能实验室","sort_order":0},{"forum_id":4058,"role":"chair","name":"曹亚男","affiliation":"中国科学院信息工程研究所","sort_order":0},{"forum_id":4058,"role":"guest","name":"蔺琛皓","affiliation":"西安交通大学","sort_order":1},{"forum_id":4058,"role":"guest","name":"朱天清","affiliation":"澳门城市大学","sort_order":2},{"forum_id":4058,"role":"guest","name":"王维强","affiliation":"蚂蚁集团","sort_order":3}],"host":"","image_url":"/api/forums/4058/cover"},{"id":4078,"title":"Second Data Intelligence Forum: Building the Data Foundation for Agents","topic":"Data Science and Data Spaces","kind":"专题论坛","chair":"童咏昕","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城2号馆-204","description":"The forum will focus on the data foundation for agents, with emphasis on data management systems for agents, the organization and governance of agent data, and new data infrastructure in the agent era.\nThe forum will discuss the following questions: What kind of data management systems do agents need to build? How can heterogeneous data such as knowledge, tools, memory, and interaction traces be managed uniformly to support continuous learning and collaboration of agents? How can agent data quality be evaluated and improved, and high-quality task data and feedback data be constructed?\nThe reason for recommendation lies in its frontier nature and importance: as artificial intelligence moves from \"large models\" to \"agents,\" the data foundation is expanding from static corpora to dynamic environments and interaction data. Building a data foundation that supports agent perception, decision-making, execution, and continuous evolution has become a key support for agent development and an important direction of common concern for academia and industry.","is_demo":0,"cover_url":"/api/forums/4078/cover?size=thumb","cover_alt":"Second Data Intelligence Forum: Building the Data Foundation for Agents","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4078","source_checked_at":"2026-10-30","agenda_json":"[[\"Agent for Data: From Data Preparation to Analysis\",\"高云君\"],[\"Data Science Systems in the Agent Era\",\"王健楠\"],[\"Efficient Construction of Large Model Agent Systems\",\"赵鑫\"],[\"Agentic Data Scientist: Building Data Science Agents for the Open World\",\"刘浩\"],[\"The Evolution of Big Data Architecture in the Agent Era\",\"李钰\"]]","code":"CNCC2026-4078","forum_type":"专题论坛","format":null,"theme_code":"数据科学与数据空间","theme_name":"数据科学与数据空间","venue":"西博城2号馆","room":"西博城2号馆-204","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4078","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第二届数据智能论坛——构筑智能体的数据基础","original_topic":"数据科学与数据空间","original_description":"论坛将聚焦智能体的数据基础，重点探讨面向智能体的数据管理系统、智能体数据的组织与治理、以及智能体时代的新型数据基础设施。\n论坛将围绕以下问题展开讨论：智能体需要构建什么样的数据管理系统？如何统一管理知识、工具、记忆、交互轨迹等异构数据，并支撑智能体的持续学习与协同？如何评估和提升智能体数据质量，构建高质量任务数据和反馈数据？\n推荐原因在于其前沿性和重要性：随着人工智能从“大模型”迈向“智能体”，数据基础正从静态语料扩展到动态环境与交互数据。构建支撑智能体感知、决策、执行和持续进化的数据基础，已成为智能体发展的关键支撑，也是学术界与产业界共同关注的重要方向。","people":[{"forum_id":4078,"role":"chair","name":"童咏昕","affiliation":"北京航空航天大学","sort_order":0},{"forum_id":4078,"role":"guest","name":"高云君","affiliation":"浙江大学","sort_order":0},{"forum_id":4078,"role":"guest","name":"王宇翔","affiliation":"北京航空航天大学","sort_order":1},{"forum_id":4078,"role":"guest","name":"王健楠","affiliation":"清华大学","sort_order":2},{"forum_id":4078,"role":"guest","name":"赵鑫","affiliation":"中国人民大学","sort_order":3},{"forum_id":4078,"role":"guest","name":"刘浩","affiliation":"香港科技大学（广州）","sort_order":4},{"forum_id":4078,"role":"guest","name":"李钰","affiliation":"阿里云计算有限公司","sort_order":5}],"host":"","image_url":"/api/forums/4078/cover"},{"id":4083,"title":"The 3rd Forum on Frontier Technologies and Applications of Compilation Systems: Compiler Evolution from Hardware-Software Co-Design to Cognitive Intelligence","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"赵捷","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-404","description":"Compilation systems are a key hub connecting algorithms and hardware. The 3rd Forum on Frontier Technologies and Applications of Compilation Systems takes \"compiler evolution from hardware-software co-design to cognitive intelligence\" as its main thread, specially inviting a number of renowned scholars to conduct in-depth discussions on frontier directions such as hardware-software co-design, AI-driven compiler optimization, accelerator virtualization, automatic code generation, and the deep integration of programming languages and AI. The forum is committed to building a high-level exchange platform for academia and industry, promoting theoretical innovation and industrial implementation of compilation systems in the era of intelligent computing. Compared with the second edition, which focused on compilation transformation in the era of large models, this edition expands its vision to broader intelligent computing scenarios; while continuing the main thread of AI compilation, it newly incorporates topics such as hardware-software co-design, system virtualization, repository-level code optimization, and cognitive intelligence, covering a full-stack perspective from architecture to upper-layer applications and achieving a comprehensive upgrade from \"thematic focus\" to \"domain convergence.\"","is_demo":0,"cover_url":"/api/forums/4083/cover?size=thumb","cover_alt":"The 3rd Forum on Frontier Technologies and Applications of Compilation Systems: Compiler Evolution from Hardware-Software Co-Design to Cognitive Intelligence","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4083","source_checked_at":"2026-10-30","agenda_json":"[[\"From Syntax Paradigms to Cognitive Intelligence: The Deep Integration and Evolution of Compilers, Programming Languages, IDEs, and AI Agents\",\"高耀清\"],[\"CLoopGen: An Automatic Executable Program Construction Framework for Loop Optimization Models\",\"黄春\"],[\"Accelerator Virtualization and Hybrid Deployment Technologies for High Utilization\",\"陈全\"],[\"MoKOpt: A Skill-Mixing Framework for Repository-Level Code Optimization\",\"姚培森\"],[\"Deterministic Spatiotemporal Computing AI Systems via Architecture-Compiler Co-Design\",\"钱学海\"]]","code":"CNCC2026-4083","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-404","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4083","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第三届编译系统前沿技术与应用——从软硬协同到认知智能的编译进化","original_topic":"新型计算与网络架构","original_description":"编译系统是连接算法与硬件的关键枢纽。第三届编译系统前沿技术与应用论坛以“从软硬协同到认知智能的编译进化”为主线，特邀多位知名学者，围绕软硬件协同设计、AI驱动编译优化、加速器虚拟化、代码自动生成及编程语言与AI深度融合等前沿方向展开深度研讨。论坛致力于为学术界与工业界搭建高水平交流平台，推动编译系统在智能计算时代的理论创新与产业落地。相较于第二届聚焦大模型时代的编译变革，本届将视野拓展至更广阔的智能计算场景，在延续AI编译主线的同时，新纳入软硬协同设计、系统虚拟化、仓库级代码优化与认知智能等议题，覆盖从体系结构到上层应用的全栈视角，实现从“专题聚焦”到“领域交汇”的全面升级。","people":[{"forum_id":4083,"role":"guest","name":"高耀清","affiliation":"华为公司","sort_order":0},{"forum_id":4083,"role":"chair","name":"赵捷","affiliation":"湖南大学","sort_order":0},{"forum_id":4083,"role":"guest","name":"黄春","affiliation":"国防科技大学计算机学院","sort_order":1},{"forum_id":4083,"role":"guest","name":"陈全","affiliation":"上海交通大学","sort_order":2},{"forum_id":4083,"role":"guest","name":"姚培森","affiliation":"浙江大学","sort_order":3},{"forum_id":4083,"role":"guest","name":"钱学海","affiliation":"清华大学","sort_order":4}],"host":"","image_url":"/api/forums/4083/cover"},{"id":4084,"title":"The 3rd \"Large-Small Model Collaborative Intelligent Computing\" Forum: From Cloud Large Models to Device-Cloud Integration","topic":"AI Foundation Models","kind":"专题论坛","chair":"王田","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-512","description":"This thematic forum is part of the \"Large-Small Model Collaborative Intelligent Computing\" series (two sessions have been successfully held, each with enthusiastic response). This session focuses on \"the path from cloud large models to device-cloud integration,\" and addresses challenges in large-model deployment such as high cost, long latency, privacy sensitivity, and device heterogeneity, discussing a new paradigm of collaborative evolution among cloud large models, device-side small models, and edge nodes. The forum will center on topics such as device-cloud integrated architectures, task-aware model routing and efficient inference, knowledge distillation and personalized adaptation, and secure, trustworthy, and governable computing, inviting experts from academia and industry to exchange key technologies, system practices, and application paths, and to promote large and small models from single-point capabilities to collaborative intelligence.","is_demo":0,"cover_url":"/api/forums/4084/cover?size=thumb","cover_alt":"The 3rd \"Large-Small Model Collaborative Intelligent Computing\" Forum: From Cloud Large Models to Device-Cloud Integration","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4084","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Welcome Address for the 3rd \\\"Large-Small Model Collaborative Intelligent Computing\\\"\",\"王田\"],[\"Theory and Methods of Large-Small Model Collaborative Multimodal Cognitive Computing\",\"王士进\"],[\"From Individual Intelligence to Swarm Interconnection: Evolution and Reflections on the Agent Internet\",\"姚海鹏\"],[\"Exploring Performance and Power Optimization for Large Model Training\",\"田臣\"],[\"Key Technologies of Collective Communication for Large Model Training\",\"翟恩南\"],[\"Model Collaboration and Interconnection\",\"李哲涛\"]]","code":"CNCC2026-4084","forum_type":"专题论坛","format":null,"theme_code":"AI基础模型","theme_name":"AI基础模型","venue":"西博城5号馆","room":"西博城5号馆-512","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4084","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第三届“大小模型协同智能计算”论坛：从云端大模型走向端云一体","original_topic":"AI基础模型","original_description":"本专题论坛是“大小模型协同智能计算”的系列论坛（已经成功举办两届，每次反响很热烈）。本次聚焦“从云端大模型走向端云一体的路径”，面向大模型落地中成本高、时延长、隐私敏感与设备异构等挑战，探讨云端大模型、端侧小模型与边缘节点协同演进的新范式。论坛将围绕端云一体架构、任务感知的模型路由与高效推理、知识蒸馏与个性化适配、安全可信与可治理计算等议题，邀请学术界和产业界专家交流关键技术、系统实践和应用路径，推动大小模型从单点能力走向协同智能。","people":[{"forum_id":4084,"role":"chair","name":"王田","affiliation":"北京师范大学","sort_order":0},{"forum_id":4084,"role":"guest","name":"王士进","affiliation":"科大讯飞股份有限公司","sort_order":0},{"forum_id":4084,"role":"guest","name":"姚海鹏","affiliation":"北京邮电大学","sort_order":1},{"forum_id":4084,"role":"guest","name":"田臣","affiliation":"南京大学","sort_order":2},{"forum_id":4084,"role":"guest","name":"翟恩南","affiliation":"阿里云计算有限公司","sort_order":3},{"forum_id":4084,"role":"guest","name":"李哲涛","affiliation":"暨南大学","sort_order":4},{"forum_id":4084,"role":"guest","name":"曾建电","affiliation":"北京师范大学","sort_order":5},{"forum_id":4084,"role":"guest","name":"秦慧玲","affiliation":"北京师范大学","sort_order":6},{"forum_id":4084,"role":"guest","name":"张程","affiliation":"北京师范大学","sort_order":7},{"forum_id":4084,"role":"guest","name":"梁玉珠","affiliation":"北京师范大学","sort_order":8},{"forum_id":4084,"role":"guest","name":"黄凤怡","affiliation":"北京师范大学","sort_order":9}],"host":"","image_url":"/api/forums/4084/cover"},{"id":4085,"title":"The Sensing-Computing Technology Revolution Driven by Optoelectronic Integration","topic":"Unclassified","kind":"专题论坛","chair":"董星辰","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城2号馆-202","description":"Existing edge sensing systems follow a discrete von Neumann architecture of \"optical imaging-photoelectric conversion-pixel buffering-electronic processor computation.\" Under the extreme constraints on massive data processing, bandwidth, and energy consumption in the AI era, the separated sensing-computing architecture faces enormous challenges. Optoelectronic integrated sensing-computing technology returns to the optical signal itself and uses the parallelism and low energy consumption of photons at the sensing source to complete most intelligent computation, becoming a foundational transformative route for breaking through the bottlenecks of electronic computing power, power consumption, and bandwidth in the post-Moore era. It will reconstruct vision and multimodal sensing systems and is a core underlying supporting technology for edge AI, embodied intelligence, and next-generation vision foundation models. This forum will focus on optoelectronic integrated sensing-computing technology, discussing academic frontier issues in key technical areas such as novel optoelectronic integrated sensing-computing architectures, optical metasurface feature extraction, all-optical reconfigurable parallel computing, and spectral intelligent sensing and reconstruction, providing a platform for exchange and learning for peers engaged in perception, cognition, and computing.","is_demo":0,"cover_url":"/api/forums/4085/cover?size=thumb","cover_alt":"The Sensing-Computing Technology Revolution Driven by Optoelectronic Integration","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4085","source_checked_at":"2026-10-30","agenda_json":"[[\"Research Progress on Mid-Infrared Single-Photon Detection and Photon Counting\",\"林春\"],[\"From Computational Imaging to Computational Visual Cognition\",\"袁鑫\"],[\"Research on 3D Scene Object Perception Based on Optoelectronic Hybrid Neural Networks\",\"邹承均\"],[\"Scalable Physical Optimization Based on Photonic-Spintronic Hybrid Computing\",\"曲俞睿\"],[\"High-Speed Optoelectronic Co-Packaged Switching Technology for Large-Scale Computing Centers\",\"杜清扬\"]]","code":"CNCC2026-4085","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城2号馆","room":"西博城2号馆-202","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4085","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"光电融合驱动下的感算技术革命","original_topic":"未分类","original_description":"现有边缘感知系统遵循“光成像-光电转换-像素缓存-电子处理器运算”的分立冯诺依曼架构，在AI时代海量数据处理及带宽能耗极限约束下，感算分离架构面临巨大挑战。光电融合感算技术，回归光信号本身，在感知源头利用光子并行、低能耗特性完成大部分智能计算，成为后摩尔时代突破电子算力、功耗、带宽瓶颈的底层变革路线，将重构视觉与多模态感知体系，是边缘AI、具身智能、下一代视觉大模型的核心底层支撑技术。本论坛将聚焦光电融合感算一体技术，从新型光电融合感算架构、光学超表面特征提取、全光可重构并行计算、光谱智能感知与重建等关键技术领域探讨学术前沿问题，为从事感知认知与计算的同行提供交流和学习的平台。","people":[{"forum_id":4085,"role":"chair","name":"董星辰","affiliation":"启元实验室","sort_order":0},{"forum_id":4085,"role":"guest","name":"林春","affiliation":"中国科学院上海技术物理研究所、红外探测全国重点实验室","sort_order":0},{"forum_id":4085,"role":"guest","name":"袁鑫","affiliation":"西湖大学","sort_order":1},{"forum_id":4085,"role":"guest","name":"邹承均","affiliation":"中国科学院微电子研究所","sort_order":2},{"forum_id":4085,"role":"guest","name":"曲俞睿","affiliation":"上海科技大学","sort_order":3},{"forum_id":4085,"role":"guest","name":"杜清扬","affiliation":"之江实验室","sort_order":4}],"host":"","image_url":"/api/forums/4085/cover"},{"id":4095,"title":"3D Perception, World Modeling and Interaction for Complex Scenes","topic":"Unclassified","kind":"专题论坛","chair":"李崇寿","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-414","description":"3D perception, world modeling and interaction are important foundations for intelligent systems to understand the physical environment and act autonomously. As applications expand from controlled scenes to open, dynamic and even extreme environments, integrating geometric structure, semantic information and spatial memory to build understandable, generatable and interactive world models has become the key to connecting visual perception with planning and decision-making and to advancing spatial intelligence and embodied intelligence. Currently, 3D perception in complex scenes still faces challenges such as incomplete observations, dynamic interference and limited sensing, making it difficult to balance accuracy and robustness; there is a lack of effective coordination among scene representation, spatial memory and content generation, and models still fall short in continuous updating, controllable editing and prediction of action consequences. These problems constrain agents' reliable reasoning, long-horizon task execution and generalization to real environments. This forum will hold discussions around 3D reconstruction and multi-source localization in complex environments, hierarchical spatial memory, explicit 3D scene representation and controllable generation, interactive world models, and multimodal agent planning and decision-making, focusing on how to connect perception, modeling and action, improve model understandability, operability and interaction reliability, and, in conjunction with applications such as aerospace exploration, visual navigation for the blind and robotics, exchange common challenges and technical paths.","is_demo":0,"cover_url":"/api/forums/4095/cover?size=thumb","cover_alt":"3D Perception, World Modeling and Interaction for Complex Scenes","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4095","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Ceremony\",\"李崇寿\"],[\"Visual Understanding and Spatial Perception for Vision-Based Blind Navigation Systems\",\"山世光\"],[\"Intelligent 3D Vision in Extreme Aerospace Environments\",\"戴玉超\"],[\"Multimodal Agents: Frontier Research and Development Trends\",\"雷印杰\"],[\"3D Scene Generation for Spatial Intelligence\",\"郭裕兰\"],[\"Toward Interactive Spatial World Models: Explicit 3D Scene Understanding, Representation, and Editing\",\"李欣科\"]]","code":"CNCC2026-4095","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-414","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4095","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"面向复杂场景的三维感知、世界建模与交互","original_topic":"未分类","original_description":"三维感知、世界建模与交互是智能系统理解物理环境并自主行动的重要基础。随着应用从受控场景拓展至开放、动态乃至极端环境，融合几何结构、语义信息与空间记忆，构建可理解、可生成、可交互的世界模型，成为连接视觉感知与规划决策、推动空间智能和具身智能发展的关键。当前，复杂场景下的三维感知仍面临观测不完备、动态干扰与传感受限等挑战，精度与鲁棒性难以兼顾；场景表征、空间记忆与内容生成之间缺乏有效协同，模型的持续更新、可控编辑及对行动后果的预测能力仍有不足。这些问题制约了智能体的可靠推理、长期任务执行与真实环境泛化。本论坛将围绕复杂环境三维重建与多源定位、层次化空间记忆、显式三维场景表征与可控生成、可交互世界模型及多模态智能体规划决策展开讨论，重点探讨如何贯通感知、建模与行动，提升模型的可理解性、可操作性与交互可靠性，并结合空天探测、视觉导盲和机器人等应用，交流共性挑战与技术路径。","people":[{"forum_id":4095,"role":"chair","name":"李崇寿","affiliation":"西南交通大学","sort_order":0},{"forum_id":4095,"role":"guest","name":"山世光","affiliation":"中国科学院计算技术研究所","sort_order":0},{"forum_id":4095,"role":"guest","name":"戴玉超","affiliation":"西北工业大学电子信息学院","sort_order":1},{"forum_id":4095,"role":"guest","name":"雷印杰","affiliation":"四川大学","sort_order":2},{"forum_id":4095,"role":"guest","name":"郭裕兰","affiliation":"中山大学","sort_order":3},{"forum_id":4095,"role":"guest","name":"李欣科","affiliation":"City Univeristy of Hong Kong","sort_order":4}],"host":"","image_url":"/api/forums/4095/cover"},{"id":4103,"title":"From Prediction to Discovery: Spatiotemporal AI Driving a New Paradigm for Complex Systems Research","topic":"Unclassified","kind":"专题论坛","chair":"王飞","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城2号馆-207","description":"Spatiotemporal complex systems widely exist in natural, social, and engineering domains, and their evolution relates to important scientific and engineering problems such as the Earth's environment, urban operations, ocean evolution, energy security, and public health. They are important research objects of AI4Science for scientific discovery and complex decision-making. In recent years, around directions such as spatiotemporal foundation models, multi-task pretraining, benchmarking, and reinforcement learning, spatiotemporal artificial intelligence is gradually forming shared foundational capabilities for multiple domains, driving complex systems research to expand from single prediction tasks toward unified modeling, complex reasoning, and autonomous decision-making. However, for real scientific problems, spatiotemporal AI still faces key challenges such as insufficient adaptability of foundational network architectures, limited cross-task generalization ability of pretrained models, difficulty in establishing a unified and fair Benchmark system, and insufficient integration of physical laws, domain knowledge, and artificial intelligence models. With the theme \"From Prediction to Discovery: Spatiotemporal AI Driving a New Paradigm for Complex Systems Research,\" this forum will discuss the above issues and focus on typical scenarios such as weather prediction, ocean eddy analysis, smart cities, and complex network optimization, exploring the deep integration of general artificial intelligence methods and domain knowledge, and promoting complex systems research from prediction toward cognition, decision-making, and scientific discovery. We will invite experts and scholars from the fields of artificial intelligence, data science, and AI4Science to jointly discuss the key theories, core technologies, and future development directions of spatiotemporal AI-driven complex systems research, and to promote interdisciplinary integration and innovative cooperation. This topic closely follows the development frontier of AI4Science, responds to key challenges in complex systems research such as insufficient transformation of predictive capability into scientific discovery and limited knowledge extraction despite abundant data, and has both theoretical innovation value and major application significance. We look forward to building a high-level academic exchange platform for CNCC and promoting the development of a new paradigm for spatiotemporal AI-driven complex systems scientific research.","is_demo":0,"cover_url":"/api/forums/4103/cover?size=thumb","cover_alt":"From Prediction to Discovery: Spatiotemporal AI Driving a New Paradigm for Complex Systems Research","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4103","source_checked_at":"2026-10-30","agenda_json":"[[\"Spatiotemporal AI-Driven Scientific Research on Complex Systems: Status and Challenges\",\"王飞\"],[\"Spatiotemporal Multi-Task Foundation Models: General Generalization and Ocean Intelligence in Practice\",\"邵泽志\"],[\"Reinforcement Learning-Driven Spatiotemporal Dynamic Computing and Complex Reasoning\",\"黄高\"],[\"OpenTS: Temporal Intelligence and Autonomous Decision-Making\",\"杨彬\"],[\"Meteorological Large Models: Theory and Practice of Physical Intelligence\",\"龙明盛\"],[\"AI for Combinatorial Problems in Complex Networks\",\"范长俊\"]]","code":"CNCC2026-4103","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城2号馆","room":"西博城2号馆-207","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4103","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"从预测到发现：时空AI驱动复杂系统科研新范式","original_topic":"未分类","original_description":"时空复杂系统广泛存在于自然、社会和工程领域，其演化关系到地球环境、城市运行、海洋演变、能源安全、公共健康等重要科学与工程问题，是AI4Science面向科学发现与复杂决策的重要研究对象。近年来，围绕时空基础模型、多任务预训练、基准评测、强化学习等方向，时空人工智能正逐步形成面向多领域共享的基础能力，推动复杂系统研究由单一预测任务向统一建模、复杂推理和自主决策拓展。然而，面向真实科学问题，时空AI仍面临基础网络架构适配性不足，预训练模型跨任务泛化能力有限，统一公平的Benchmark体系难建立，以及物理规律、领域知识与人工智能模型融合不充分等关键挑战。本论坛以“从预测到发现：时空AI驱动复杂系统科研新范式”为主题，将围绕上述议题展开讨论，并聚焦气象预测、海洋涡旋分析、智慧城市与复杂网络优化等典型场景，探讨通用人工智能方法与领域知识的深度融合，推动复杂系统研究由预测走向认知、决策与科学发现。我们将邀请来自人工智能、数据科学及AI4Science领域的专家学者，共同探讨时空AI驱动复杂系统研究的关键理论、核心技术及未来发展方向，促进跨学科交叉融合与创新合作。该选题紧扣AI4Science的发展前沿，回应复杂系统研究中预测能力向科学发现转化不足、数据丰富而知识提炼有限等关键挑战，兼具理论创新价值与重大应用意义。期待为CNCC搭建高水平的学术交流平台，推动时空AI驱动复杂系统科学研究新范式的发展。","people":[{"forum_id":4103,"role":"chair","name":"王飞","affiliation":"中国科学院计算技术研究所","sort_order":0},{"forum_id":4103,"role":"guest","name":"邵泽志","affiliation":"中国科学院计算技术研究所","sort_order":0},{"forum_id":4103,"role":"guest","name":"黄高","affiliation":"清华大学","sort_order":1},{"forum_id":4103,"role":"guest","name":"杨彬","affiliation":"华东师范大学","sort_order":2},{"forum_id":4103,"role":"guest","name":"龙明盛","affiliation":"清华大学","sort_order":3},{"forum_id":4103,"role":"guest","name":"范长俊","affiliation":"国防科学技术大学","sort_order":4}],"host":"","image_url":"/api/forums/4103/cover"},{"id":4104,"title":"Breaking the Compute Barrier for On-Device Intelligence: The Path to Mass Production of Compute-in-Memory Chips and the Device Ecosystem","topic":"Unclassified","kind":"专题论坛","chair":"全书学、刘方鑫","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-107B三星堆B","description":"AI large models are accelerating from the cloud to the device: smartphones, AI PCs, wearable devices, and embodied intelligence terminals all require running tens-of-billions-parameter models locally in real time. However, the power budget of on-device devices is often only a few watts, and under the traditional von Neumann architecture, data movement energy consumption already dominates absolutely; the \"memory wall\" and \"power wall\" are more acute on the device side than in data centers. Compute-in-memory, by embedding computation into memory arrays, fundamentally eliminates data movement overhead and achieves energy efficiency comparable to advanced process nodes using mature processes; it is a key path for on-device AI chips to achieve \"leapfrog development\" and has also been listed as a frontier architecture direction in the national \"15th Five-Year Plan\". This forum focuses on the technical challenges and mass-production implementation of on-device compute-in-memory, with topics including:\n• Compute-in-memory chip architecture and circuit design based on SRAM and DRAM processes: the evolution of digital and analog compute-in-memory architectures, and collaborative optimization of arrays and peripheral circuits;\n• Chip architecture design for on-device large-model inference: the generality contradiction among compute density, energy efficiency ratio, and rapid algorithm iteration; • how compilers and software stacks can be compatible with mainstream ecosystems such as Triton, enabling terminal developers to migrate without friction;\n• Mass-production pain points such as consistency of large-scale arrays, yield, and automated toolchains;\n• From the perspective of terminal manufacturers: how compute-in-memory chips can truly \"get on the device\"—collaboration with smartphone SoCs, operating systems, and application ecosystems;\n• The cost accounting and business closed loop of on-device PIM under the device-cloud collaboration landscape. The forum will bring together compute-in-memory chip companies, terminal manufacturers, and university scholars, covering the complete on-device industry chain from chips and complete machines to applications, and discussing the key leap of this technology from tape-out to large-scale implementation.","is_demo":0,"cover_url":"/api/forums/4104/cover?size=thumb","cover_alt":"Breaking the Compute Barrier for On-Device Intelligence: The Path to Mass Production of Compute-in-Memory Chips and the Device Ecosystem","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4104","source_checked_at":"2026-10-30","agenda_json":"[[\"Requirements and Challenges of Terminal Products for On-Device Large Models\",\"王卫兵\"],[\"On-Device Large Model Inference Driven by Heterogeneous Compute-in-Memory and 3D Integration\",\"刘琦\"],[\"Pioneering the 3D-CIM™ Architecture for 3D Compute-in-Memory, Opening a Trillion-Scale New Track for Large Model Inference!\",\"叶乐\"],[\"Three Laws of General-Purpose Compute-in-Memory\",\"熊大鹏\"],[\"Industrialization and Application of On-Device Large Model AI Chips\",\"卞大志\"],[\"Tensor Systolic, Four-Level Leap: Metax's Breakthrough Practice in On-Device Large Model Chips\",\"李凯\"],[\"Rethinking the System Position of Compute-in-Memory in the Era of On-Device Large Models\",\"蒋力\"]]","code":"CNCC2026-4104","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-107B三星堆B","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":7,"forum_id":"CNCC2026-4104","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"端侧智能的算力破局：存算一体芯片的量产之路与终端生态","original_topic":"未分类","original_description":"AI 大模型正加速从云端走向终端：手机、AI PC、可穿戴设备与具身智能终端都要求在地实时运行百亿参数级模型。然而端侧设备的功耗预算往往只有几瓦，传统冯·诺依曼架构下数据搬运能耗已占绝对主导，\"存储墙\"与\"功耗墙\"在端侧表现得比在数据中心更为尖锐。存算一体通过将计算嵌入存储阵列，从根源上消除搬运开销，以成熟工艺实现比肩先进制程的能效，是端侧 AI 芯片实现\"弯道超车\"的关键路径，也被列为国家\"十五五\"规划的前沿架构方向。本论坛聚焦端侧存算一体的技术攻坚与量产落地，议题包括：\n• 基于 SRAM 与 DRAM 工艺的存算一体芯片架构与电路设计：数字与模拟存算架构的演进、阵列与外围电路的协同优化；\n• 面向端侧大模型推理的芯片架构设计：算力密度、能效比与算法快速迭代之间的通用性矛盾；•编译器与软件栈如何兼容 Triton 等主流生态，让终端开发者无感迁移；\n• 大规模阵列的一致性、良率与自动化工具链等量产痛点；\n• 终端厂商视角：存算一体芯片如何真正\"上机\"——与手机 SoC、操作系统及应用生态的协同；\n• 端云协同格局下端侧 PIM 的成本账与商业闭环。论坛将汇聚存算一体芯片公司、终端厂商与高校学者，从芯片、整机到应用完整覆盖端侧产业链，探讨这一技术从流片走向规模化落地的关键一跃。","people":[{"forum_id":4104,"role":"chair","name":"全书学","affiliation":"亿铸科技","sort_order":0},{"forum_id":4104,"role":"guest","name":"王卫兵","affiliation":"OPPO","sort_order":0},{"forum_id":4104,"role":"co_chair","name":"刘方鑫","affiliation":"上海交通大学","sort_order":1},{"forum_id":4104,"role":"guest","name":"刘琦","affiliation":"复旦大学","sort_order":1},{"forum_id":4104,"role":"guest","name":"叶乐","affiliation":"北京大学","sort_order":2},{"forum_id":4104,"role":"guest","name":"熊大鹏","affiliation":"亿铸科技","sort_order":3},{"forum_id":4104,"role":"guest","name":"卞大志","affiliation":"光羽芯辰","sort_order":4},{"forum_id":4104,"role":"guest","name":"李凯","affiliation":"迈特芯","sort_order":5},{"forum_id":4104,"role":"guest","name":"蒋力","affiliation":"上海交通大学","sort_order":6}],"host":"","image_url":"/api/forums/4104/cover"},{"id":4109,"title":"Technology Stacking or Paradigm Reconstruction? Opportunities and Paths for Web 3.0 in the AI Era","topic":"General","kind":"专题论坛","chair":"王伟","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-503","description":"Web 3.0 is a new generation of the Internet built on blockchain as its technological foundation. Unlike Web 1.0 and Web 2.0, Web 3.0 enables users to autonomously control their digital identities and data assets, promoting data rights confirmation and value circulation. Currently, AI technologies such as large models and AI Agents are driving Web 3.0 from automatic contract fulfillment via smart contracts toward intelligent decision-making and autonomous collaboration, and the blueprint for a next-generation digital economy of human-machine collaborative co-governance is accelerating its formation. At this window of opportunity, as the 15th Five-Year Plan's national key R&D program lays out its blockchain special project, the integration of AI and Web 3.0 also raises important questions that urgently require in-depth discussion. Is Web 3.0 in the AI era merely technology stacking on the existing system, or is it a paradigm reconstruction? How will AI reshape the circulation of Web 3.0 digital assets? What unprecedented opportunities and challenges lie beneath this integration? The forum will focus on these questions, inviting outstanding experts and scholars in related fields to report on the latest achievements of Web 3.0, and in the discussion session to jointly explore the opportunities and paths for Web 3.0 in the AI era.","is_demo":0,"cover_url":"/api/forums/4109/cover?size=thumb","cover_alt":"Technology Stacking or Paradigm Reconstruction? Opportunities and Paths for Web 3.0 in the AI Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4109","source_checked_at":"2026-10-30","agenda_json":"[[\"Post-Quantum Blockchain Systems: Current Status and Prospects\",\"斯雪明\"],[\"Web 3.0 in the Agent Economy: Architectural Evolution and Technical Challenges\",\"成秀珍\"],[\"Blockchain-Based Trustworthy Token Metering for Agents\",\"朱建明\"],[\"Exploration of Web3 Multi-Chain Interconnection Technologies\",\"孙毅\"],[\"Agent Payment and PayFi: Payment Infrastructure for the Machine Economy\",\"刘涛\"]]","code":"CNCC2026-4109","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城5号馆","room":"西博城5号馆-503","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4109","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"技术叠加还是范式重构？AI时代Web 3.0的机遇与路径","original_topic":"综合","original_description":"Web 3.0 是以区块链为技术底座的新一代互联网。区别于Web 1.0和Web 2.0，Web 3.0可实现用户对数字身份和数据资产的自主控制，推动数据确权与价值流通。当前，大模型与智能体等AI技术正推动Web 3.0从智能合约自动履约迈向智能决策与自主协作，人机协同共治的下一代数字经济蓝图正加速形成。值此“十五五”国家重点研发计划区块链专项布局的窗口期，AI 与 Web 3.0 的融合也提出了亟待深度探讨的重要问题。AI 时代的 Web 3.0，究竟只是既有体系的技术叠加，还是范式重构？AI将如何重塑Web 3.0数字资产流通方式？融合之下又潜藏哪些前所未有的机遇与挑战？论坛将聚焦这些问题，邀请相关杰出专家学者报告 Web 3.0的最新成果，并在研讨环节共同探讨 AI 时代 Web 3.0 的机遇与路径。","people":[{"forum_id":4109,"role":"chair","name":"王伟","affiliation":"西安交通大学","sort_order":0},{"forum_id":4109,"role":"guest","name":"斯雪明","affiliation":"上海交通大学","sort_order":0},{"forum_id":4109,"role":"guest","name":"李超","affiliation":"北京交通大学","sort_order":1},{"forum_id":4109,"role":"guest","name":"成秀珍","affiliation":"山东大学","sort_order":2},{"forum_id":4109,"role":"guest","name":"朱建明","affiliation":"中央财经大学","sort_order":3},{"forum_id":4109,"role":"guest","name":"孙毅","affiliation":"中国科学院计算技术研究所","sort_order":4},{"forum_id":4109,"role":"guest","name":"刘涛","affiliation":"北京世纪互联宽带数据中心有限公司","sort_order":5}],"host":"","image_url":"/api/forums/4109/cover"},{"id":4124,"title":"Tenth Knowledge Graph Forum—Knowledge Enhancement and Logical Governance: Knowledge Graphs Empowering Trustworthy AI","topic":"AI Foundation Models","kind":"专题论坛","chair":"陈华钧","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-406兴隆湖","description":"With the rapid development of large language models and agent technologies, knowledge enhancement and logical governance are becoming key paths to breaking through model hallucinations, improving complex reasoning capabilities, and building trustworthy AI. As an important infrastructure connecting data, knowledge, and intelligence, knowledge graphs are deeply integrating with dynamic ontologies, GraphRAG, neuro-symbolic learning, knowledge hypergraphs, and other technologies, driving the continuous evolution of knowledge representation, semantic understanding, logical reasoning, and autonomous decision-making capabilities. With the theme \"Knowledge Enhancement and Logical Governance: Knowledge Graphs Empowering Trustworthy AI,\" this forum invites renowned experts in related fields to share the latest theories, key technologies, and application practices around frontier directions such as the integration of dynamic ontologies and large models, knowledge-enhanced reasoning, graph agents, neuro-symbolic learning, and knowledge hypergraphs, and to jointly discuss the development trends of knowledge graphs empowering trustworthy AI, providing new ideas for building trustworthy, interpretable, and controllable intelligent systems.","is_demo":0,"cover_url":"/api/forums/4124/cover?size=thumb","cover_alt":"Tenth Knowledge Graph Forum—Knowledge Enhancement and Logical Governance: Knowledge Graphs Empowering Trustworthy AI","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4124","source_checked_at":"2026-10-30","agenda_json":"[[\"Neuro-Symbolic Learning\",\"杨博\"],[\"Exploration of Reasoning and Decision-Making Paradigms Based on Knowledge Hypergraphs\",\"田玲\"],[\"Research on Integrated Fusion Technology of Dynamic Ontology and Large Models for Complex Adversarial Environments\",\"姜巍\"],[\"Data Representation and Knowledge Enhancement of Large Language Models\",\"张日崇\"],[\"From Graph Agents to Autonomous Discovery Systems: Structured Knowledge Empowering Trustworthy Scientific Reasoning\",\"鄂海红\"]]","code":"CNCC2026-4124","forum_type":"专题论坛","format":null,"theme_code":"AI基础模型","theme_name":"AI基础模型","venue":"天府国际会议中心","room":"天府国际会议中心-406兴隆湖","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4124","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第十届知识图谱论坛——知识增强与逻辑驾驭：知识图谱赋能可信AI","original_topic":"AI基础模型","original_description":"随着大语言模型和智能体技术的快速发展，知识增强与逻辑驾驭正成为突破模型幻觉、提升复杂推理能力、构建可信AI的关键路径。知识图谱作为连接数据、知识与智能的重要基础设施，正与动态本体、GraphRAG、神经符号学习、知识超图等技术深度融合，推动知识表示、语义理解、逻辑推理和自主决策能力持续演进。本论坛以“知识增强与逻辑驾驭：知识图谱赋能可信AI”为主题，邀请相关领域知名专家，围绕动态本体与大模型融合、知识增强推理、图谱智能体、神经符号学习、知识超图等前沿方向，分享最新理论、关键技术与应用实践，共同探讨知识图谱赋能可信AI的发展趋势，为构建可信、可解释、可控的智能系统提供新思路。","people":[{"forum_id":4124,"role":"chair","name":"陈华钧","affiliation":"浙江大学计算机科学与技术学院","sort_order":0},{"forum_id":4124,"role":"guest","name":"杨博","affiliation":"吉林大学","sort_order":0},{"forum_id":4124,"role":"guest","name":"田玲","affiliation":"电子科技大学","sort_order":1},{"forum_id":4124,"role":"guest","name":"姜巍","affiliation":"中国电科智能科技研究院","sort_order":2},{"forum_id":4124,"role":"guest","name":"张日崇","affiliation":"北京航空航天大学","sort_order":3},{"forum_id":4124,"role":"guest","name":"鄂海红","affiliation":"北京邮电大学","sort_order":4},{"forum_id":4124,"role":"guest","name":"林俊宇","affiliation":"复旦大学","sort_order":5}],"host":"","image_url":"/api/forums/4124/cover"},{"id":4130,"title":"Device-Cloud Integration-Driven Efficient Content Production: Opportunities, Challenges, and Practice","topic":"AI for Industry","kind":"专题论坛","chair":"吴飞","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-511","description":"This forum focuses on device-cloud integration-driven efficient content generation, and around directions such as generative content safety governance, social media behavior simulation, multimodal content generation, and scientific intelligence integration, discusses key bottlenecks and breakthrough paths in technological evolution. Current content generation technology is facing a structural contradiction between continuously improving model capabilities and device-side resource constraints, and the safety and trustworthiness of generated content and the efficiency of cross-modal collaboration have become core issues constraining large-scale application. The forum focuses on risk regulation mechanisms for generative content, cognitive communication modeling of social media behavior, and efficient collaborative generation strategies for multimodal content, exploring optimization paths for computing power allocation and data collaboration under device-cloud integrated architectures. It aims to promote content generation from a single modality to cross-modality, and from centralized processing to device-cloud collaboration, advancing technological innovation and application deployment of device-cloud integration technology in the content production field.","is_demo":0,"cover_url":"/api/forums/4130/cover?size=thumb","cover_alt":"Device-Cloud Integration-Driven Efficient Content Production: Opportunities, Challenges, and Practice","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4130","source_checked_at":"2026-10-30","agenda_json":"[[\"A Cognitive Communication-Driven Social Media Simulator\",\"鲍秉坤\"],[\"Knowledge-Fusion AIGC Research for AI for Science\",\"颜波\"],[\"Safety Cognition Construction and Knowledge Reshaping for Generative Multimedia Models\",\"谢洪涛\"],[\"From Image Generation to 3D Generation\",\"高联丽\"],[\"Efficient and Trustworthy Cross-Modal Content Generation\",\"赵洲\"]]","code":"CNCC2026-4130","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"西博城5号馆","room":"西博城5号馆-511","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4130","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"端云融合驱动的高效内容生产：机遇、挑战与实践","original_topic":"AI+行业","original_description":"本论坛聚焦端云融合驱动的高效内容生成，围绕生成式内容安全治理、社交媒体行为模拟、多模态内容生成与科学智能融合等方向，探讨技术演进中的关键瓶颈与突破路径。当前内容生成技术正面临模型能力持续提升与端侧资源约束之间的结构性矛盾，生成内容的安全可信与跨模态协同效率已成为制约规模化应用的核心问题。论坛重点关注生成式内容的风险调控机制、社交媒体行为的认知传播建模以及多模态内容的高效协同生成策略，探索端云融合架构下算力分配与数据协同的优化路径。旨在推动内容生成从单一模态向跨模态拓展，从中心化处理向端云协同演进，促进端云融合技术在内容生产领域的技术创新与应用落地。","people":[{"forum_id":4130,"role":"chair","name":"吴飞","affiliation":"浙江大学","sort_order":0},{"forum_id":4130,"role":"guest","name":"鲍秉坤","affiliation":"合肥工业大学","sort_order":0},{"forum_id":4130,"role":"guest","name":"颜波","affiliation":"复旦大学","sort_order":1},{"forum_id":4130,"role":"guest","name":"谢洪涛","affiliation":"中国科学技术大学","sort_order":2},{"forum_id":4130,"role":"guest","name":"高联丽","affiliation":"电子科技大学","sort_order":3},{"forum_id":4130,"role":"guest","name":"赵洲","affiliation":"浙江大学","sort_order":4}],"host":"","image_url":"/api/forums/4130/cover"},{"id":4148,"title":"Token Operations: From Scale Production and Intelligent Scheduling to Value Monetization","topic":"General","kind":"专题论坛","chair":"吴杰","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-515","description":"Tokens are the core computational unit and value carrier in the era of large models, and have now become an important link connecting computing infrastructure, scheduling systems, and business models. The concept of \"Token operations\" advocates systematic operations around the full lifecycle of tokens—from efficient production at the底层 (bottom layer), through intelligent distribution in the middle, to value monetization at the top—to build a new technology-economy system for large model services.\n\nThis forum is organized around three major sections: token scale production, intelligent scheduling, and value monetization, and discusses topics related to high-quality data engineering, model inference acceleration, distributed computing scheduling, token pricing, and the business ecosystem. The conference invites experts in system architecture, algorithms, networking, and economics to discuss frontier topics such as data infrastructure, efficient inference frameworks, computing scheduling solutions, token economics, and implementation practices, building an interdisciplinary exchange platform between academia and industry to facilitate the exchange of ideas and industrial innovation.","is_demo":0,"cover_url":"/api/forums/4148/cover?size=thumb","cover_alt":"Token Operations: From Scale Production and Intelligent Scheduling to Value Monetization","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4148","source_checked_at":"2026-10-30","agenda_json":"[[\"The Palantir Phenomenon and the Development of China's Data Industry\",\"周傲英\"],[\"AI Ping: A One-Stop Token Service Platform\",\"翟季冬\"],[\"Distributed Edge Computing Power Networks for High-Quality Large Model Agent Services: Theory, Technologies, and Practical Cases\",\"王晓飞\"],[\"Token Economics: Pricing and Mechanism Design in the Era of Large Models\",\"郑臻哲\"],[\"Practice of the Xingchen TokenHub Operations Service Platform\",\"杨东明\"]]","code":"CNCC2026-4148","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城5号馆","room":"西博城5号馆-515","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4148","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"Token经营：从规模生产、智能调度到价值变现","original_topic":"综合","original_description":"Token 是大模型时代核心计算单元与价值载体，现已成为串联算力设施、调度系统和商业模式的重要纽带。\"Token经营\"理念主张围绕Token全生命周期进行系统化运营——从底层高效生产、中间智能分发到上层价值变现，构建大模型服务的新型技术-经济体系。\n\n本次论坛围绕 Token 规模生产、智能调度、价值变现三大板块展开，探讨高质量数据工程、模型推理加速、分布式算力调度、Token 定价与商业生态相关内容。会议邀约系统架构、算法、网络、经济学领域专家，研讨数据基础设施、高效推理框架、算力调度方案、Token经济学及落地实践等前沿话题，搭建学界和产业界跨学科交流平台，助力思想互通与产业创新。","people":[{"forum_id":4148,"role":"guest","name":"周傲英","affiliation":"华东师范大学","sort_order":0},{"forum_id":4148,"role":"chair","name":"吴杰","affiliation":"中国电信云计算研究院","sort_order":0},{"forum_id":4148,"role":"guest","name":"翟季冬","affiliation":"清华大学","sort_order":1},{"forum_id":4148,"role":"guest","name":"王晓飞","affiliation":"天津大学","sort_order":2},{"forum_id":4148,"role":"guest","name":"郑臻哲","affiliation":"上海交通大学","sort_order":3},{"forum_id":4148,"role":"guest","name":"杨东明","affiliation":"天翼云科技有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4148/cover"},{"id":4154,"title":"How Agents Are Reshaping Enterprise Services: From Technical Breakthroughs to Practical Implementation","topic":"Unclassified","kind":"专题论坛","chair":"王昊奋、王玉翔","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-411","description":"In recent years, large model applications represented by enterprise service agents are driving profound changes in service forms—evolving from traditional rule engines and retrieval-based question answering toward agents with knowledge enhancement, tool invocation, and autonomous decision-making, and have become key support for industries such as finance, e-commerce, and government affairs to improve service efficiency and user experience.\nHowever, current enterprise service agents still face core challenges such as factual hallucination, long-horizon task planning, tool invocation accuracy, output controllability, and business compliance. How to promote agents from \"usable\" to \"easy to use\" urgently requires the establishment of scientific evaluation systems and trustworthy implementation frameworks. This forum focuses on the frontier technologies, evaluation, and application practices of enterprise service agents, and invites experts from academia and industry to engage in in-depth exchanges on key issues such as retrieval-augmented generation, multi-agent collaboration, and output safety guardrails, jointly discussing the implementation bottlenecks and breakthrough paths of enterprise service agent technology.","is_demo":0,"cover_url":"/api/forums/4154/cover?size=thumb","cover_alt":"How Agents Are Reshaping Enterprise Services: From Technical Breakthroughs to Practical Implementation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4154","source_checked_at":"2026-10-30","agenda_json":"[[\"Toward Autonomous Agents: The Evolution of Enterprise Services from \\\"Can Talk\\\" to \\\"Can Do\\\"\",\"黄萱菁\"],[\"Beyond the Model: What Is Still Missing for Enterprise Deployment of Voice Dialogue Agents?\",\"王春平\"],[\"Agent Security and Enterprise Governance in Practice: From \\\"Saying the Wrong Thing\\\" to \\\"Doing the Wrong Thing\\\"\",\"张谧\"],[\"Bidirectional Empowerment of Knowledge and Intelligence: A New Paradigm for Large-Model Agents\",\"雷文强\"],[\"FinLifeBench: A Cross-Stage Benchmark for Evaluating Financial Dialogue Agents\",\"高峰\"]]","code":"CNCC2026-4154","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-411","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4154","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智能体如何重塑企业服务：从技术突破到落地实战","original_topic":"未分类","original_description":"近年来，以企业服务智能体为代表的大模型应用正推动服务形态发生深刻变革——从传统规则引擎、检索式问答向知识增强、工具调用、自主决策的智能体演进，已成为金融、电商、政务等行业提升服务效率与用户体验的关键支撑。\n然而，当前企业服务智能体仍面临事实幻觉、长程任务规划、工具调用准确性、输出可控性与业务合规性等核心挑战，如何推动智能体从“可用”走向“好用”，亟需建立科学的评测体系与可信落地框架。本论坛聚焦企业服务智能体的前沿技术、评测与应用实践，邀请学界与产业界专家就检索增强生成、多智能体协同、输出安全护栏等关键问题展开深度交流，共同探讨企业服务智能体技术的落地瓶颈与突破路径。","people":[{"forum_id":4154,"role":"chair","name":"王昊奋","affiliation":"同济大学","sort_order":0},{"forum_id":4154,"role":"guest","name":"黄萱菁","affiliation":"复旦大学计算机学院","sort_order":0},{"forum_id":4154,"role":"co_chair","name":"王玉翔","affiliation":"信也科技集团","sort_order":1},{"forum_id":4154,"role":"guest","name":"王春平","affiliation":"信也科技集团","sort_order":1},{"forum_id":4154,"role":"guest","name":"张谧","affiliation":"复旦大学","sort_order":2},{"forum_id":4154,"role":"guest","name":"杨洋","affiliation":"浙江大学","sort_order":3},{"forum_id":4154,"role":"guest","name":"高峰","affiliation":"信也科技集团","sort_order":4},{"forum_id":4154,"role":"guest","name":"肖斌","affiliation":"重庆邮电大学","sort_order":5},{"forum_id":4154,"role":"guest","name":"雷文强","affiliation":"四川大学","sort_order":6},{"forum_id":4154,"role":"guest","name":"孙逸飞","affiliation":"浙江大学","sort_order":7}],"host":"","image_url":"/api/forums/4154/cover"},{"id":4162,"title":"Debating the Circulation and Use of Public Medical Data: Security vs. Convenience, Universal Benefit vs. Revenue, Availability vs. Good Use","topic":"Unclassified","kind":"专题论坛","chair":"吴伶、陈俊颖","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-204A蜀都厅A","description":"Background: High-quality medical datasets are an important foundation for medical artificial intelligence R&D and validation, clinical diagnosis and treatment improvement, and public health governance. Currently, various regions are advancing the coordinated aggregation and authorized operation of public data in healthcare, medical insurance, and pharmaceuticals, exploring utilization models such as trusted data spaces, privacy-preserving computation, and data sandboxes. However, public medical data still face the problem of being \"aggregated but hard to use, available but not well usable\": overly cautious management may lead to long approval cycles, high usage costs, and insufficient value release; one-sided pursuit of convenience may breach the boundaries of personal information protection, medical ethics, and data security.\n\nTo break this dilemma, it is necessary both to clarify classification and grading, scenario authorization, fee allocation, and responsibility boundaries through institutions, and to transform institutional requirements into executable, verifiable, and traceable governance capabilities through technology. Trusted data spaces can support identity authentication, purpose control, and full-process auditing; privacy-preserving computation can conduct joint analysis without directly exposing raw data; data sandboxes can implement controlled trials over users, algorithmic tools, and result export. But technology is not omnipotent; it is also necessary to answer questions such as how technical routes adapt to different scenarios, who certifies technical capabilities, and who supervises platform operation. Institutions set boundaries for technology, technology supports the implementation of institutions, and new risks and needs emerging in application in turn drive the continuous improvement of governance rules.\n\nPoints for deliberation:\n\n1. Security vs. convenience—how to move from \"keeping it under control\" to \"making it usable\"?\n\nHow can classification and grading and scenario-based authorization be implemented according to data sensitivity, purpose of use, and potential risk, and translated into technical rules executable by the platform? How can trusted data spaces, privacy-preserving computation, and data sandboxes be used to prevent unauthorized use, model inversion, and information leakage? How should security strength, data precision, computational efficiency, and usage cost be balanced? Technical platforms integrate authorization, operation, and auditing, so how should they be supervised?\n\n2. Universal benefit vs. revenue—how can charging and revenue distribution be implemented?\n\nWho can prioritize the use of public medical data, in what manner and at what price, and who shares the resulting revenue? How can free or low-cost use for public-interest purposes and reasonable charging for industrial purposes be implemented? Can usage metering, digital contracts, data lineage, and process attestation provide a basis for cost accounting, charging pricing, and revenue distribution? How can excessively high technical and compliance costs be prevented from excluding small and medium-sized institutions and research teams?\n\n3. Availability vs. good use—how to move from \"making it usable\" to \"making good use of it\"?\n\nHow can a full-lifecycle governance mechanism covering data entry, controlled use, output of results, effect evaluation, and responsibility tracing be established? Can data sandboxes simultaneously serve functions of algorithm validation, ethical review, bias detection, and regulatory experimentation? How can it be evaluated whether privacy-preserving technologies truly achieve \"data available but not visible\"? When data quality problems, platform loss of control, or algorithm misuse occur, how can data lineage, operation logs, and algorithm records be used to clarify the responsibilities of all parties, while loosening constraints for compliant innovation and holding the security bottom line?","is_demo":0,"cover_url":"/api/forums/4162/cover?size=thumb","cover_alt":"Debating the Circulation and Use of Public Medical Data: Security vs. Convenience, Universal Benefit vs. Revenue, Availability vs. Good Use","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4162","source_checked_at":"2026-10-30","agenda_json":"[[\"AI and Blockchain-Empowered Secure Data Sharing\",\"张彦\"],[\"From Data Circulation to Health Creation: A New Paradigm of Public Healthcare Data Governance for Health Outcomes\",\"石昕\"],[\"Healthcare Data Governance, Intelligent Parsing, and Data-Element Value Practices\",\"刘喜恩\"],[\"Design and Progress of Health and Medical Data Assetization Pathways\",\"杨帆\"],[\"Healthcare Data Security and Value Release Based on Confidential Intelligent Computing\",\"张昊天\"]]","code":"CNCC2026-4162","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-204A蜀都厅A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4162","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"公共医疗数据流通利用之辨：安全vs.便捷，普惠vs.收益，可用vs.善用","original_topic":"未分类","original_description":"背景：高质量医疗数据集是医疗人工智能研发验证、临床诊疗改进和公共卫生治理的重要基础。当前，各地正推进医疗、医保、医药公共数据协同汇聚和授权运营，探索可信数据空间、隐私保护计算、数据沙箱等利用模式。然而，公共医疗数据仍面临“汇而难用、可用不好用”：管理过于审慎，可能导致审批周期长、使用成本高和价值释放不足；片面追求便捷，又可能突破个人信息保护、医学伦理和数据安全边界。\n\n破解这一困境，既要以制度明确分类分级、场景授权、收费分配和责任边界，也要以技术将制度要求转化为可执行、可验证、可追溯的治理能力。可信数据空间可支撑身份认证、用途控制和全程审计；隐私保护计算可在原始数据不直接暴露的条件下开展联合分析；数据沙箱可对使用主体、算法工具和结果导出实施受控试验。但技术并非万能，还需回答技术路线如何适配不同场景、技术能力由谁认证、平台运行由谁监督等问题。制度为技术划定边界，技术支撑制度落地，应用中出现的新风险和新需求又推动治理规则持续完善。\n\n思辨点：\n\n1.安全vs.便捷——如何从“管得住”走向“用得了”？\n\n如何根据数据敏感程度、使用目的和潜在风险实施分类分级与分场景授权，并将其转化为平台可执行的技术规则？如何运用可信数据空间、隐私保护计算和数据沙箱防范越权使用、模型反演和信息泄露？安全强度、数据精度、计算效率和使用成本应如何平衡？技术平台集授权、运行和审计于一体，又应如何受到监督？\n\n2.普惠vs.收益——收费和收益分配如何落地？\n\n谁可以优先使用公共医疗数据，以何种方式和价格使用，产生的收益由谁分享？如何落实公益用途免费或低成本使用、产业用途合理收费？能否通过用量计量、数字合约、数据血缘和过程存证，为成本核算、收费定价及收益分配提供依据？如何防止技术和合规成本过高，将中小机构与科研团队排除在外？\n\n3.可用vs.善用——如何从“用得了”走向“用得好”？\n\n如何建立覆盖数据进入、受控使用、成果输出、效果评价和责任追溯的全生命周期治理机制？数据沙箱能否兼具算法验证、伦理审查、偏差检测和监管试验功能？如何评价隐私保护技术是否真正实现“数据可用不可见”？发生数据质量问题、平台失控或算法滥用时，如何借助数据血缘、操作日志和算法记录明确各方责任，在为合规创新松绑的同时守住安全底线？","people":[{"forum_id":4162,"role":"chair","name":"吴伶","affiliation":"福州大学","sort_order":0},{"forum_id":4162,"role":"guest","name":"张彦","affiliation":"电子科技大学","sort_order":0},{"forum_id":4162,"role":"co_chair","name":"陈俊颖","affiliation":"华南理工大学","sort_order":1},{"forum_id":4162,"role":"guest","name":"石昕","affiliation":"中国医科大学","sort_order":1},{"forum_id":4162,"role":"guest","name":"刘喜恩","affiliation":"清华大学","sort_order":2},{"forum_id":4162,"role":"guest","name":"杨帆","affiliation":"山东大学","sort_order":3},{"forum_id":4162,"role":"guest","name":"张昊天","affiliation":"广州芳禾数据有限公司","sort_order":4},{"forum_id":4162,"role":"guest","name":"陈超","affiliation":"重庆大学","sort_order":5},{"forum_id":4162,"role":"guest","name":"王吴越","affiliation":"深绎未来科技（广东横琴）有限公司","sort_order":6}],"host":"","image_url":"/api/forums/4162/cover"},{"id":4163,"title":"AutoResearch: New Frameworks and Technologies for Autonomous Research Agents","topic":"AI Agents","kind":"专题论坛","chair":"严睿","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-410","description":"As large models demonstrate increasingly strong capabilities, an exciting vision is rapidly fermenting: can people build a full-process autonomous research system—from reading literature, designing algorithms, and completing experiments, to iterating ideas, optimizing model architectures, and even writing a complete paper? This is no longer a distant idea in science fiction narratives. Recently, autonomous research agents have rapidly developed from an experimental concept into multiple runnable prototype systems. With the theme \"New Frameworks and Technologies for Autonomous Research Agents,\" this forum will discuss research issues in directions such as personalized research, dynamic planning and reasoning, and academic innovation paradigms, and explore basic framework design and key technical breakthroughs. The forum has invited four representative speakers and multiple participating guests to give in-depth reports and discussions from four dimensions: system-level automation, personalized memory and evolution, planning and reasoning, and the academic ecosystem. The conclusions will jointly point to the ultimate purpose of automated research: to build research infrastructure capable of continuous learning and deep adaptation.","is_demo":0,"cover_url":"/api/forums/4163/cover?size=thumb","cover_alt":"AutoResearch: New Frameworks and Technologies for Autonomous Research Agents","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4163","source_checked_at":"2026-10-30","agenda_json":"[[\"AI-Empowered Research Automation\",\"张岳\"],[\"SciencePal 2.0: A Self-Evolving Personal Research Partner\",\"Ning Miao\"],[\"From General to Personalized: A Planning-Capable Automated Research Agent\",\"陈秀颖\"],[\"From AI4S to AI Scientists\",\"李勇\"]]","code":"CNCC2026-4163","forum_type":"专题论坛","format":null,"theme_code":"Agent智能体","theme_name":"Agent智能体","venue":"西博城4号馆","room":"西博城4号馆-410","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4163","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AutoResearch: 自主科研智能体新框架与新技术","original_topic":"Agent智能体","original_description":"当大模型展现越来越强的能力时，一个激动人心的愿景正迅速发酵：人们能否构建一个全流程的自主科研系统：从阅读文献、设计算法、完成实验，到迭代思路、优化模型架构，直至撰写完整论文？这已不再是科幻叙事中的遥远设想。近期，自主科研智能体从一个实验性概念迅速发展为多个可运行的原型系统。本论坛将以“自主科研智能体新框架与新技术”为主题，研讨个性化科研、动态规划推理、以及学术创新范式等方向的研究问题，探讨基础框架设计与关键技术突破。论坛邀请到四位代表性讲者与多位与会嘉宾，分别从系统级自动化、个性化记忆与进化、规划推理和学术生态四个维度展开深度报告与研讨。结论将共同指向自动科研终极目的，是构建一种能够持续学习、深度适配的科研基建。","people":[{"forum_id":4163,"role":"chair","name":"严睿","affiliation":"武汉大学","sort_order":0},{"forum_id":4163,"role":"guest","name":"张岳","affiliation":"西湖大学","sort_order":0},{"forum_id":4163,"role":"guest","name":"Ning Miao","affiliation":"香港城市大学","sort_order":1},{"forum_id":4163,"role":"guest","name":"陈秀颖","affiliation":"穆罕默德·本·扎耶德人工智能大学（阿联酋人工智能大学）","sort_order":2},{"forum_id":4163,"role":"guest","name":"杜博","affiliation":"武汉大学","sort_order":3},{"forum_id":4163,"role":"guest","name":"李勇","affiliation":"清华大学","sort_order":4},{"forum_id":4163,"role":"guest","name":"高莘","affiliation":"电子科技大学","sort_order":5},{"forum_id":4163,"role":"guest","name":"郑翔","affiliation":"香港城市大学","sort_order":6},{"forum_id":4163,"role":"guest","name":"赵迎功","affiliation":"香港城市大学","sort_order":7},{"forum_id":4163,"role":"guest","name":"闫峻","affiliation":"香港城市大学","sort_order":8}],"host":"","image_url":"/api/forums/4163/cover"},{"id":4170,"title":"2nd Underwater Unmanned Swarm Intelligent Collaboration Technology Forum","topic":"Robotics and Embodied Intelligence","kind":"专题论坛","chair":"於志文","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-409","description":"Underwater multi-agent collaboration integrates multidisciplinary technologies such as underwater communication, swarm cooperative control, and marine environment perception, and is a key track for multidisciplinary innovation in intelligence, control, marine engineering, and other fields. However, in complex marine environments there are inherent constraints such as limited underwater acoustic communication and ocean current disturbances, making it difficult for traditional algorithms to directly adapt to underwater scenarios; the industry has long faced multiple bottlenecks in theoretical models, networking communication, collaborative decision-making, and engineering implementation, while the digitalization and intelligent upgrading of the ocean also bring entirely new technical opportunities and industrial space for underwater swarm intelligence. This forum focuses on the full-chain technology of underwater multi-agent collaboration, gathers top frontline experts in fields such as domestic swarm intelligence computing, underwater robots, and underwater acoustic communication, builds a cross-boundary dialogue platform among academia, industry, and research institutes, breaks down the barriers between theoretical algorithms and marine engineering applications, builds consensus on the development of the underwater swarm intelligence field, and provides high-level exchange and cooperation opportunities for young scholars and technical R&D personnel.","is_demo":0,"cover_url":"/api/forums/4170/cover?size=thumb","cover_alt":"2nd Underwater Unmanned Swarm Intelligent Collaboration Technology Forum","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4170","source_checked_at":"2026-10-30","agenda_json":"[[\"Subsea AUV and its case studies\",\"陈鹰\"],[\"From Deep Sea to Ultra-Shallow Water: Cross-Scenario Application Challenges and Technological Innovation Pathways for Marine Robots\",\"俞建成\"],[\"Bionic Crowd Intelligence Sensing and Computing\",\"郭斌\"],[\"Research on Cooperative Control Technology for Underwater Unmanned Swarms in Formation Navigation\",\"徐健\"],[\"Probabilistic Forecasting of Stochastic Dynamical Systems: Ceilings, Floors, and the Space in Between\",\"何建平\"]]","code":"CNCC2026-4170","forum_type":"专题论坛","format":null,"theme_code":"机器人与具身智能","theme_name":"机器人与具身智能","venue":"西博城4号馆","room":"西博城4号馆-409","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4170","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第二届水下无人群体智能协同技术论坛","original_topic":"机器人与具身智能","original_description":"水下多智能体协同融合水下通信、集群协同控制、海洋环境感知等多学科交叉技术，是智能、控制、海工等多学科交叉创新的关键赛道。但复杂海洋环境下存在水声通信受限、海流扰动等天然约束，传统算法难以直接适配水下场景，行业长期面临理论模型、组网通信、协同决策、工程落地多重瓶颈，同时海洋数字化、智能化升级也为水下集群智能带来全新技术机遇与产业空间。本论坛聚焦水下多智能体协同全链条技术，汇聚国内群智计算、水下机器人、水声通信等领域一线顶尖专家，搭建学术界+工业界+科研院所跨界对话平台，打通理论算法与海洋工程应用壁垒，凝聚水下集群智能领域发展共识，为青年学者、技术研发人员提供高水平交流与合作契机。","people":[{"forum_id":4170,"role":"chair","name":"於志文","affiliation":"哈尔滨工程大学","sort_order":0},{"forum_id":4170,"role":"guest","name":"陈鹰","affiliation":"浙江大学","sort_order":0},{"forum_id":4170,"role":"guest","name":"俞建成","affiliation":"厦门大学","sort_order":1},{"forum_id":4170,"role":"guest","name":"郭斌","affiliation":"西北工业大学","sort_order":2},{"forum_id":4170,"role":"guest","name":"徐健","affiliation":"哈尔滨工程大学","sort_order":3},{"forum_id":4170,"role":"guest","name":"何建平","affiliation":"上海交通大学","sort_order":4}],"host":"","image_url":"/api/forums/4170/cover"},{"id":4172,"title":"Can AI Truly Understand and Interact with the Physical World?","topic":"Vision, Multimodality and World Models","kind":"专题论坛","chair":"杨鑫、徐凯","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-402","description":"With the deep integration of artificial intelligence in scientific research and engineering practice, physical interaction intelligence is becoming an important paradigm driving next-generation technological innovation. It aims to closely combine the mechanisms of the physical world with data-driven intelligent methods to achieve closed-loop optimization from virtual simulation to physical systems. This forum focuses on three key directions: AI for Science, intelligent simulation, and embodied interaction, and discusses how to advance the discovery of scientific laws, accelerate physical modeling, and improve the autonomous intelligence level of physical systems through artificial intelligence. The forum aims to distill core scientific questions and key technical bottlenecks in the field, discuss open problems and future development directions, and form a conclusion or consensus report, providing reference and guidance for theoretical breakthroughs and industrial applications of physical interaction intelligence.","is_demo":0,"cover_url":"/api/forums/4172/cover?size=thumb","cover_alt":"Can AI Truly Understand and Interact with the Physical World?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4172","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Remarks\",\"徐凯\"],[\"Forum Introduction\",\"杨鑫\"],[\"Research on Object Perception Technology for Complex Scenarios\",\"张建国\"],[\"Physically Grounded Generative World Modeling\",\"陈志波\"],[\"Exploring New Large Model Architectures: JTok and xHC\",\"严骏驰\"],[\"Two Decades of Progress in Computational Imaging Reconstruction Algorithms: From Iterative Optimization to Deep Learning and Generative AI\",\"袁鑫\"],[\"Brain-Inspired Online Learning Technology for Device-Side \\\"Learning-Inference\\\" Integration: From \\\"Static Inference\\\" to \\\"Dynamic Intelligence\\\"\",\"文飞\"]]","code":"CNCC2026-4172","forum_type":"专题论坛","format":null,"theme_code":"视觉、多模态与世界模型","theme_name":"视觉、多模态与世界模型","venue":"西博城4号馆","room":"西博城4号馆-402","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":7,"forum_id":"CNCC2026-4172","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI能否真正理解物理世界并与之交互？","original_topic":"视觉、多模态与世界模型","original_description":"随着人工智能在科学研究与工程实践中的深度融合，物理交互智能正成为推动下一代技术创新的重要范式。它旨在将物理世界的机理与数据驱动的智能方法紧密结合，实现从虚拟仿真到实体系统的闭环优化。本次论坛聚焦三大关键方向：科学智能、智能仿真与具身交互，探讨如何通过人工智能推进科学规律发现、加速物理建模、并提升实体系统的自主智能水平。论坛旨在凝练领域核心科学问题与关键技术瓶颈，探讨开放性问题与未来发展方向，并形成结论或共识报告，为物理交互智能的理论突破与产业应用提供参考与指引。","people":[{"forum_id":4172,"role":"chair","name":"杨鑫","affiliation":"大连理工大学","sort_order":0},{"forum_id":4172,"role":"guest","name":"张建国","affiliation":"南方科技大学","sort_order":0},{"forum_id":4172,"role":"co_chair","name":"徐凯","affiliation":"中国科学院工业人工智能研究所","sort_order":1},{"forum_id":4172,"role":"guest","name":"陈志波","affiliation":"中国科学技术大学","sort_order":1},{"forum_id":4172,"role":"guest","name":"严骏驰","affiliation":"上海交通大学","sort_order":2},{"forum_id":4172,"role":"guest","name":"袁鑫","affiliation":"西湖大学","sort_order":3},{"forum_id":4172,"role":"guest","name":"文飞","affiliation":"上海交通大学","sort_order":4}],"host":"","image_url":"/api/forums/4172/cover"},{"id":4173,"title":"AI-Empowered Cybersecurity: Innovative Teaching and Research Practices and Talent Cultivation Exploration","topic":"Unclassified","kind":"专题论坛","chair":"严明","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-506","description":"This forum differs from conventional smart education and general AI teaching forums in that its core feature is taking cybersecurity as the underlying main line. Generative AI is changing the landscape of cybersecurity offense and defense, posing entirely new challenges to cybersecurity discipline teaching and research and talent cultivation. This forum focuses on \"AI-Empowered Cybersecurity: Innovative Teaching and Research Practices and Talent Cultivation Exploration,\" bringing together university teaching and research experts and industry technical representatives to exchange and discuss frontier AI security technologies, AI security + teaching and research, AI security + curriculum system reform, practical teaching development, and industry-education collaborative talent cultivation.","is_demo":0,"cover_url":"/api/forums/4173/cover?size=thumb","cover_alt":"AI-Empowered Cybersecurity: Innovative Teaching and Research Practices and Talent Cultivation Exploration","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4173","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Remarks\",\"严明\"],[\"Challenges and Countermeasures in AI Model Security Construction from a Practical Perspective\",\"于锐\"],[\"Large Model Security Offense and Defense: From Frontier Research to Practical Applications and Cybersecurity Talent Cultivation\",\"杜彦辉\"],[\"AI-Empowered Cybersecurity Talent Cultivation: Problems, Concepts, Practices, and Expected Outcomes\",\"袁斌\"],[\"Future-Oriented Security Operations Technology Practices\",\"张振礼\"],[\"Course Release: Large Model Security and Application Technologies\",\"王洪军\"]]","code":"CNCC2026-4173","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-506","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4173","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI 赋能网络安全：教研创新实践与人才培养探索","original_topic":"未分类","original_description":"本论坛区别于常规智慧教育、通用 AI 教学类论坛，核心特色在于以网络安全为底层主线，生成式 AI 正改变网络安全攻防格局，对网安学科教研与人才培养提出全新挑战。本论坛聚焦“AI 赋能网络安全：教研创新实践与人才培养探索”，汇聚高校教研专家、产业技术代表，围绕AI安全前沿技术、AI安全+教研、AI安全+课程体系改革、实践教学建设、产教协同育人开展交流研讨。","people":[{"forum_id":4173,"role":"chair","name":"严明","affiliation":"公安部科信委","sort_order":0},{"forum_id":4173,"role":"guest","name":"王洪军","affiliation":"深信服科技股份有限公司","sort_order":0},{"forum_id":4173,"role":"guest","name":"于锐","affiliation":"公安部第一研究所","sort_order":1},{"forum_id":4173,"role":"guest","name":"杜彦辉","affiliation":"中国人民公安大学","sort_order":2},{"forum_id":4173,"role":"guest","name":"袁斌","affiliation":"华中科技大学","sort_order":3},{"forum_id":4173,"role":"guest","name":"张振礼","affiliation":"深信服科技股份有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4173/cover"},{"id":4177,"title":"Opportunities, Challenges, and Integrated Development of Women in Science and Technology in the Era of Large-Scale Research Collaboration","topic":"Unclassified","kind":"专题论坛","chair":"刘家瑛、金一","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-412","description":"When research shifts from \"individual wisdom + small teams\" to \"large-scale computing collaboration + open science + global team networks,\" is this an \"equalization revolution\" that breaks old barriers, or a \"computing divide\" that exacerbates resource stratification? In this paradigm change, are women in science and technology being pushed to a more favorable position, or toward a more marginalized zone?\nThe forum has designed two points for deliberation:\nPoint for deliberation one: Capability spectrum—in the intelligent computing era, are women's research strengths being \"amplified\" or \"diluted\"?\nSpecifically: traits traditionally regarded as women's strengths, such as meticulousness, collaboration, and interdisciplinary integration, in computing-intensive, engineering-oriented intelligent computing research—do they become more valuable (e.g., data governance, model interpretability, ethical alignment), or are they systematically marginalized (with attention only to engineering implementation speed and resource mobilization capability)?\nPoint for deliberation two: Path reconstruction—how can women in science and technology \"find their own coordinates\" in the computing-restructured landscape?\nSpecifically: under the new infrastructure situation of \"large-scale computing collaboration + open science + global team networks,\" how should women in science and technology find their own coordinates and direction (strive for computing resources, enter core tracks, or redefine research questions and open new fields with lower dependence on computing)? How should women in science and technology pursue integrated development?","is_demo":0,"cover_url":"/api/forums/4177/cover?size=thumb","cover_alt":"Opportunities, Challenges, and Integrated Development of Women in Science and Technology in the Era of Large-Scale Research Collaboration","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4177","source_checked_at":"2026-10-30","agenda_json":"[[\"Context Engineering Optimization for Complex Marketing Decision-Making\",\"杜军平\"],[\"Reflections on Large Model Technology\",\"吴华\"],[\"Research and Application of Embodied Intelligent Surgical Navigation for Endoscopy\",\"宋红\"],[\"Toward Emergent Intelligence in the Physical World\",\"仉尚航\"],[\"Breaking Out and Establishing Oneself: Women in Science and Technology at Industry-Featured Universities—A Path of Distinctive Development in the Era of Interdisciplinary Integration\",\"王杨\"]]","code":"CNCC2026-4177","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-412","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4177","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"大规模科研协作时代女科技工作者的机遇、挑战与融合发展","original_topic":"未分类","original_description":"当科研从\"个人智慧+小团队\"转向\"大规模算力协作+开放科学+全球团队网络\"，这究竟是一场打破旧壁垒的\"平权革命\"，还是一场加剧资源分化的\"算力鸿沟\"？女科技工作者在这场范式变革中，是被推到了更有利的位置，还是被推向了更边缘的地带？\n论坛设计了两个思辨点：\n思辨点一：能力光谱——智算时代，女性的科研优势是被\"放大\"还是被\"稀释\"？\n具体： 传统认知中女性擅长的细致、协作、跨学科整合等特质，在算力密集型、工程导向的智算科研中，是变得更有价值（如数据治理、模型可解释性、伦理对齐），还是被系统性边缘化（只看工程实现速度和资源调动能力）？\n思辨点二：路径重构——女科技工作者如何在算力重构的版图中\"找准自己的坐标\"？\n具体： \"大规模算力协作+开放科学+全球团队网络\"新基建形势下，女科技工作者应该如何找准自己的坐标和方向（争取算力资源、进入核心赛道还是重新定义研究问题、开辟算力依赖度更低的新领域）？女科技工作者应当如何融合发展？","people":[{"forum_id":4177,"role":"chair","name":"刘家瑛","affiliation":"北京大学","sort_order":0},{"forum_id":4177,"role":"guest","name":"王杨","affiliation":"西南石油大学计算机科学学院","sort_order":0},{"forum_id":4177,"role":"co_chair","name":"金一","affiliation":"北京交通大学","sort_order":1},{"forum_id":4177,"role":"guest","name":"杜军平","affiliation":"北京邮电大学","sort_order":1},{"forum_id":4177,"role":"guest","name":"吴华","affiliation":"百度","sort_order":2},{"forum_id":4177,"role":"guest","name":"宋红","affiliation":"北京理工大学","sort_order":3},{"forum_id":4177,"role":"guest","name":"仉尚航","affiliation":"北京大学","sort_order":4},{"forum_id":4177,"role":"guest","name":"王鹏","affiliation":"北京中关村学院","sort_order":5},{"forum_id":4177,"role":"guest","name":"李群","affiliation":"北京嘀嘀无限科技发展有限公司","sort_order":6}],"host":"","image_url":"/api/forums/4177/cover"},{"id":4179,"title":"4th Human-Centered Intelligent Recommender Systems Forum — Intent-Driven Recommender Systems: A New Paradigm of Human-AI Collaboration in the Agent Era","topic":"AI Foundation Models","kind":"专题论坛","chair":"卢暾","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-408","description":"This forum is the fourth Human-Centered Intelligent Recommender Systems forum that Fudan University has held consecutively at CNCC. Continuing the value stance of \"human-centeredness,\" the ethical framework of \"responsibility,\" and the interaction paradigm of \"human-AI collaboration,\" this edition further advances the focus from \"how humans collaborate with AI\" to \"how humans express, negotiate, and execute intentions through AI,\" with the theme \"Intent-Driven Recommender Systems: A New Paradigm of Human-AI Collaboration in the Agent Era.\" The core academic concern of the forum is: when large-model-empowered agents become the intermediary layer between users and digital services, the design logic of recommender systems is undergoing a paradigm shift from \"implicit behavior inference\" to \"explicit intent negotiation,\" and the user's role shifts from a modeled object to a co-definer of recommendation objectives and constraints; its essence is a fundamental reconstruction of the control architecture in recommender systems. Around this core proposition, the forum will focus on research challenges at three levels: intent modeling and alignment, human-AI negotiated recommendation mechanisms, and governable recommender system architectures. The forum invites top scholars in fields such as recommender systems, human-computer interaction, and large model governance to conduct in-depth discussions on the above issues, striving to provide a clear theoretical framework and feasible technical paths for human-centered recommender systems in the agent era.","is_demo":0,"cover_url":"/api/forums/4179/cover?size=thumb","cover_alt":"4th Human-Centered Intelligent Recommender Systems Forum — Intent-Driven Recommender Systems: A New Paradigm of Human-AI Collaboration in the Agent Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4179","source_checked_at":"2026-10-30","agenda_json":"[[\"Immersive Perception of the Brain and Personalized Recommendation\",\"张敏\"],[\"Large Language Models and Human Values: Simulation, Control, and Bias\",\"郑雯\"],[\"Large Model-Based Simulation of Individual Human Behavior\",\"陈旭\"],[\"Professional Media Knowledge Services: From Content Recommendation to Human-AI Collaboration\",\"陈自强\"],[\"Autonomous Knowledge Management: From Manual Summarization to Automatic Accumulation\",\"雷文强\"]]","code":"CNCC2026-4179","forum_type":"专题论坛","format":null,"theme_code":"AI基础模型","theme_name":"AI基础模型","venue":"西博城4号馆","room":"西博城4号馆-408","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4179","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第四届以人为本的智能推荐系统论坛——意图驱动的推荐系统：智能体时代的人智协同新范式","original_topic":"AI基础模型","original_description":"本届论坛是复旦大学在CNCC上连续举办的第四届以人为本的智能推荐系统论坛，延续“以人为中心”的价值立场、“负责任”的伦理框架与“人智协同”的交互范式，本届进一步将焦点从“人如何与AI协作”推进至“人如何通过AI表达、协商与执行意图”，主题为“意图驱动的推荐系统：智能体时代的人智协同新范式”。论坛的核心学术关切在于：当大模型赋能的智能体成为用户与数字服务之间的中介层，推荐系统的设计逻辑正在经历从“隐式行为推断”向“显式意图协商”的范式转变，用户的角色从被建模的客体转变为推荐目标与约束条件的共同定义者，其本质是控制权架构在推荐系统中的根本性重构。围绕这一核心命题，论坛将着重探讨意图建模与对齐、人智协商式推荐机制、以及可治理推荐系统架构三个层面的研究挑战。论坛邀请推荐系统、人机交互、大模型治理等领域的顶尖学者，围绕上述问题展开深度研讨，力求为以人为本的推荐系统在智能体时代提供清晰的理论框架与可行的技术路径。","people":[{"forum_id":4179,"role":"chair","name":"卢暾","affiliation":"复旦大学","sort_order":0},{"forum_id":4179,"role":"guest","name":"李东胜","affiliation":"微软亚洲研究院","sort_order":0},{"forum_id":4179,"role":"guest","name":"张敏","affiliation":"清华大学","sort_order":1},{"forum_id":4179,"role":"guest","name":"郑雯","affiliation":"复旦大学","sort_order":2},{"forum_id":4179,"role":"guest","name":"陈旭","affiliation":"中国人民大学","sort_order":3},{"forum_id":4179,"role":"guest","name":"陈自强","affiliation":"新华网股份有限公司","sort_order":4},{"forum_id":4179,"role":"guest","name":"雷文强","affiliation":"四川大学","sort_order":5},{"forum_id":4179,"role":"guest","name":"李诺","affiliation":"上海对外经贸大学","sort_order":6}],"host":"","image_url":"/api/forums/4179/cover"},{"id":4181,"title":"Data-Intelligence Integration: Building AI-Ready High-Quality Datasets and Vertical Domain Model Innovation","topic":"Data Science and Data Spaces","kind":"专题论坛","chair":"丁晓锋","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-510","description":"As vertical large models accelerate their large-scale deployment, data produced by traditional data processing models generally suffers from problems such as semantic fragmentation and insufficient adaptability, making it difficult to support model fine-tuning, retrieval augmentation, and scenario reasoning, and the supply gap for industry-oriented AI-Ready high-quality datasets continues to widen.\n\nBased on the development trend of data-intelligence integration, this forum brings together experts in data governance, databases, and artificial intelligence to discuss the construction of high-quality dataset systems, focusing on key technologies such as multi-source heterogeneous data fusion, dynamic quality verification, and domain sample purification, and directly addressing pain points such as data dispersion, high labeling costs, and distribution imbalance.\n\nThe forum advocates establishing AI-native data quality evaluation standards around dimensions such as semantic alignment, sample purity, and distribution balance, and promoting quality verification to shift from post-delivery data cleaning to rule embedding at the labeling stage, providing reusable and traceable solutions for vertical domain datasets and supporting continuous iteration of large models and deep industry deployment.","is_demo":0,"cover_url":"/api/forums/4181/cover?size=thumb","cover_alt":"Data-Intelligence Integration: Building AI-Ready High-Quality Datasets and Vertical Domain Model Innovation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4181","source_checked_at":"2026-10-30","agenda_json":"[[\"Forum Introduction\",\"丁晓锋\"],[\"Data-Centric AI: From High-Quality Dataset Construction to Efficient Training and Reliable Inference of Domain-Specific Models\",\"陈雷\"],[\"Structured Data Synthesis, Governance, and Domain Large Model Training\",\"伍赛\"],[\"High-Quality Graph Dataset Construction and Intelligent Applications\",\"张岩峰\"],[\"Multimodal Perception and Autonomous Decision-Making for Low-Altitude Embodied Intelligent Agents\",\"俞俊\"],[\"Industrial Time-Series Data Governance: KaiwuDB Empowering Innovation in Domain-Specific Large Models\",\"陈磊\"]]","code":"CNCC2026-4181","forum_type":"专题论坛","format":null,"theme_code":"数据科学与数据空间","theme_name":"数据科学与数据空间","venue":"西博城5号馆","room":"西博城5号馆-510","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4181","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"数智融合:面向AI‑Ready高质量数据集构建与垂域模型革新","original_topic":"数据科学与数据空间","original_description":"随着垂直大模型加快规模化落地，传统数据加工模式产出的数据普遍存在语义割裂、适配性不足等问题，难以支撑模型微调、检索增强与场景推理，面向行业的AI-Ready高质量数据集供给缺口持续扩大。\n\n本论坛立足数智融合发展趋势，汇聚数据治理、数据库与人工智能领域专家，围绕高质量数据集体系建设展开研讨，聚焦多源异构数据融合、动态质量校验、领域样本提纯等关键技术，直面数据分散、标注成本高、分布失衡等痛点。\n\n论坛倡导围绕语义对齐、样本纯度、分布均衡等维度建立AI原生数据质量评价标准，推动质量校验由数据交付后清洗转向标注阶段规则内嵌，为垂直领域数据集提供可复用、可追溯的方案，支撑大模型持续迭代与行业深度落地。","people":[{"forum_id":4181,"role":"chair","name":"丁晓锋","affiliation":"华中科技大学","sort_order":0},{"forum_id":4181,"role":"guest","name":"陈雷","affiliation":"香港科技大学","sort_order":0},{"forum_id":4181,"role":"guest","name":"伍赛","affiliation":"浙江大学","sort_order":1},{"forum_id":4181,"role":"guest","name":"张岩峰","affiliation":"东北大学","sort_order":2},{"forum_id":4181,"role":"guest","name":"俞俊","affiliation":"哈尔滨工业大学（深圳）","sort_order":3},{"forum_id":4181,"role":"guest","name":"陈磊","affiliation":"浪潮集团","sort_order":4},{"forum_id":4181,"role":"guest","name":"唐靖","affiliation":"香港科技大学（广州）","sort_order":5},{"forum_id":4181,"role":"guest","name":"苗嘉旭","affiliation":"哈尔滨工业大学（深圳）（哈尔滨工业大学深圳科技创新研究院）","sort_order":6}],"host":"","image_url":"/api/forums/4181/cover"},{"id":4184,"title":"Foundation Models in Autonomous Driving and Embodied Intelligence Scenarios","topic":"Robotics and Embodied Intelligence","kind":"专题论坛","chair":"史少帅","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-407","description":"Foundation models are reshaping the technological paradigms of autonomous driving and embodied intelligence. From end-to-end driving decision-making to physical world interaction, from VLA models to world models, foundation models not only enhance the system's ability to understand and reason about complex scenarios, but also become a key bridge connecting digital intelligence and the physical world. Autonomous driving is regarded as the first large-scale implementation scenario of physical AI, and foundation models are precisely the core engine driving this paradigm leap.\nThis forum specially invites top experts from academia and industry to share the latest research results and technical insights of foundation models in the fields of autonomous driving and embodied intelligence, both showcasing frontier breakthroughs in the field and encouraging in-depth collision of viewpoints and extraction of consensus. We look forward to this forum pointing out the direction for foundation model research in autonomous driving and embodied intelligence, and promoting intelligent systems from \"seeing the world\" to \"understanding the world\" and \"interacting with the world.\"","is_demo":0,"cover_url":"/api/forums/4184/cover?size=thumb","cover_alt":"Foundation Models in Autonomous Driving and Embodied Intelligence Scenarios","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4184","source_checked_at":"2026-10-30","agenda_json":"[[\"Intelligent Emergence: The Path of Autonomous Driving Technology Innovation and Industrialization\",\"史少帅\"],[\"Toward Next-Generation End-to-End Autonomous Driving: Frontier Exploration of Efficient and Unified Vision-Language-Action (VLA) Models\",\"李鸿升\"],[\"World-Model-Driven End-to-End Autonomous Driving\",\"蒋理\"],[\"Modeling, Generation, and Interaction for Open 3D Worlds\",\"齐晓娟\"],[\"Qwen-Drive: A Preliminary Exploration of Vision-Language Foundation Models for Autonomous Driving\",\"梁定康\"]]","code":"CNCC2026-4184","forum_type":"专题论坛","format":null,"theme_code":"机器人与具身智能","theme_name":"机器人与具身智能","venue":"西博城4号馆","room":"西博城4号馆-407","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4184","legacy_id":4,"lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"自动驾驶与具身智能场景下的基础模型","original_topic":"机器人与具身智能","original_description":"基础模型正在重塑自动驾驶与具身智能的技术范式。从端到端驾驶决策到物理世界交互，从VLA模型到世界模型，基础模型不仅提升了系统对复杂场景的理解与推理能力，更成为连接数字智能与物理世界的关键桥梁。自动驾驶被认为是物理AI的首个大规模落地场景，而基础模型正是驱动这一范式跃迁的核心引擎。\n本论坛特邀学术界与产业界的顶尖专家，分享基础模型在自动驾驶与具身智能领域的最新研究成果与技术洞察，既集中展示领域前沿突破，也鼓励深度观点碰撞与共识提炼。我们期待通过本次论坛，为自动驾驶与具身智能的基础模型研究指明方向，推动智能系统从“看见世界”走向“理解世界”与“交互世界”。","people":[{"forum_id":4184,"role":"chair","name":"史少帅","affiliation":"滴滴自动驾驶公司","sort_order":0},{"forum_id":4184,"role":"guest","name":"李鸿升","affiliation":"香港中文大学","sort_order":0},{"forum_id":4184,"role":"guest","name":"蒋理","affiliation":"香港中文大学（深圳）","sort_order":1},{"forum_id":4184,"role":"guest","name":"齐晓娟","affiliation":"香港大学","sort_order":2},{"forum_id":4184,"role":"guest","name":"梁定康","affiliation":"华中科技大学","sort_order":3}],"host":"","image_url":"/api/forums/4184/cover"},{"id":4185,"title":"AI Empowering Accessibility: New Opportunities, Challenges, and Paradigm Innovation for the Development of Persons with Disabilities from Survival Assurance to Participatory Co-construction","topic":"AI for Industry","kind":"专题论坛","chair":"袁鑫、许俊林","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-101海螺沟","description":"Introduction: At a time when AI technology is reconstructing the foundations of the digital society, the development of persons with disabilities is ushering in a critical turning point from \"survival assurance\" to \"participatory co-construction.\" Based on the theme of CNCC2026, \"Data-Driven, Intelligent Computing for the Future,\" this forum focuses on three core issues: how AI can break down barriers to social participation for persons with disabilities, reshape individual development capabilities, and build a new ecosystem of integration between persons with and without disabilities. The forum breaks away from a one-way public welfare narrative and, with perspective debate as the main thread, sets up three progressively layered topics: \"barrier removal—capability reshaping—ecosystem reconstruction.\" It brings together top accessibility AI scholars, young researchers with disabilities in the computer field, and experts in social governance and industry standards to engage in exchanges and collisions from multiple perspectives such as technology implementation, research participation, and co-building the ecosystem. The forum aims to clarify the realistic paths and key bottlenecks of AI empowering the development of persons with disabilities, promote a shift in social focus from \"basic survival\" to \"value creation,\" from \"passive acceptance\" to \"active co-construction,\" and explore a new paradigm for high-quality integrated development of persons with and without disabilities in the digital era.","is_demo":0,"cover_url":"/api/forums/4185/cover?size=thumb","cover_alt":"AI Empowering Accessibility: New Opportunities, Challenges, and Paradigm Innovation for the Development of Persons with Disabilities from Survival Assurance to Participatory Co-construction","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4185","source_checked_at":"2026-10-30","agenda_json":"[[\"From Support to Empowerment: Educational Practice and Information Accessibility Exploration by Researchers with Disabilities\",\"姚登峰\"],[\"A Preliminary Study of Human-Centered Wearable Visual Assistance Systems\",\"顾磊磊\"],[\"Human-Machine-Environment Integration for Empowering Information Accessibility\",\"周晟\"]]","code":"CNCC2026-4185","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"天府国际会议中心","room":"天府国际会议中心-101海螺沟","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":3,"forum_id":"CNCC2026-4185","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI赋能无障碍发展：从生存保障到参与共建的残疾人发展新机遇、新挑战与范式革新","original_topic":"AI+行业","original_description":"简介：在AI技术重构数字社会根基的当下，残疾人发展正迎来从“生存保障”迈向“参与共建”的关键拐点。本论坛立足CNCC2026“数聚驱动，智算未来”之主题，聚焦AI如何破除残障群体社会参与壁垒、重塑个体发展能力、构建残健融合新生态三大核心议题。论坛打破单向公益叙事，以观点思辨为主线，设置“壁垒破除—能力重塑—生态重构”三层递进议题，汇聚顶尖无障碍AI学者、计算机领域残障青年科研骨干、社会治理与行业标准专家，从技术落地、科研参与、共建生态等多重视角交锋碰撞。论坛旨在厘清AI赋能残疾人发展的现实路径与关键卡点，推动社会关注重点从“兜底生存”转向“价值创造”，从“被动接纳”走向“主动共建”，探索数字时代残健融合高质量发展的全新范式。","people":[{"forum_id":4185,"role":"chair","name":"袁鑫","affiliation":"武汉科技大学","sort_order":0},{"forum_id":4185,"role":"guest","name":"姚登峰","affiliation":"北京联合大学/清华大学","sort_order":0},{"forum_id":4185,"role":"co_chair","name":"许俊林","affiliation":"武汉科技大学","sort_order":1},{"forum_id":4185,"role":"guest","name":"顾磊磊","affiliation":"上海交通大学","sort_order":1},{"forum_id":4185,"role":"guest","name":"周晟","affiliation":"浙江大学","sort_order":2},{"forum_id":4185,"role":"guest","name":"李麟青","affiliation":"苏州大学计算机科学与技术学院","sort_order":3},{"forum_id":4185,"role":"guest","name":"王甦菁","affiliation":"中国科学院心理研究所","sort_order":4},{"forum_id":4185,"role":"guest","name":"矣晓沅","affiliation":"微软亚洲研究院","sort_order":5},{"forum_id":4185,"role":"guest","name":"邵镇炜","affiliation":"杭州电子科技大学","sort_order":6}],"host":"","image_url":"/api/forums/4185/cover"},{"id":4190,"title":"Anthropomorphic Interaction with Large Models: Ethics, Safety, and Value Alignment","topic":"Unclassified","kind":"专题论坛","chair":"贾珈、吴梦玥","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-508","description":"Large models are currently undergoing a profound shift from purely efficiency tools to agents with emotional companionship and anthropomorphic interaction capabilities. However, existing large-model safety standards mainly focus on \"factual and content safety\" levels such as information accuracy and refusal to output harmful content, and can hardly cover the new risks brought by anthropomorphic interaction. In light of the recently issued administrative measures for anthropomorphic interactive services by the Cyberspace Administration of China and other departments, this forum aims to fill this safety governance blind spot and is highly timely and practically significant. The forum will bring together experts from three fields—voice dialogue technology, large-model safety, and science and technology ethics—and its discussions will focus deeply on identity transparency and labeling mechanisms in ultra-realistic timbres and emotional expression, how large models can prevent inducing excessive dependence and emotional manipulation during long-term companionship, and how to transform new ethical principles into executable and measurable safety alignment metrics for dialogue systems, jointly exploring the optimal balance between technological innovation and risk prevention and control.","is_demo":0,"cover_url":"/api/forums/4190/cover?size=thumb","cover_alt":"Anthropomorphic Interaction with Large Models: Ethics, Safety, and Value Alignment","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4190","source_checked_at":"2026-10-30","agenda_json":"[[\"Affective Computing and Security in Anthropomorphic Interaction\",\"秦兵\"],[\"Multimodal Large Model Security\",\"毛震东\"],[\"Anthropomorphic Voice Interaction and Safety Alignment of Speech Large Models\",\"武执政\"],[\"Safety Alignment of Large Language Models\",\"王希廷\"],[\"Ethical Risk Assessment of Large Models in Healthcare: What Else Can We Expect Beyond Compliance?\",\"徐汉辉\"]]","code":"CNCC2026-4190","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-508","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4190","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"大模型拟人化交互：伦理、安全与价值对齐","original_topic":"未分类","original_description":"当前大模型正经历从纯粹的效率工具向具备情感陪伴、拟人化交互特性的智能体的深度转变。然而，既往的大模型安全标准主要聚焦于信息准确性、拒绝输出有害内容等“事实与内容安全”层面，难以涵盖拟人化交互带来的新型风险。结合国家网信办等部门近期出台的拟人化互动服务管理办法，本论坛旨在弥补这一安全治理盲区，具有极高的时效性与现实意义。论坛将联合语音对话技术、大模型安全、科技伦理三大领域的专家，研讨将深度聚焦于超逼真音色与情感表达中的身份透明与标识机制，大模型在长期陪伴中如何防止诱导过度依赖与情感操控，以及如何将新的伦理伦理原则转化为对话系统可执行、可测量的安全对齐指标，共同探索技术创新与风险防控的最佳平衡。","people":[{"forum_id":4190,"role":"chair","name":"贾珈","affiliation":"清华大学","sort_order":0},{"forum_id":4190,"role":"guest","name":"秦兵","affiliation":"哈尔滨工业大学计算机学院","sort_order":0},{"forum_id":4190,"role":"co_chair","name":"吴梦玥","affiliation":"上海交通大学","sort_order":1},{"forum_id":4190,"role":"guest","name":"毛震东","affiliation":"中国科学技术大学","sort_order":1},{"forum_id":4190,"role":"guest","name":"武执政","affiliation":"香港中文大学（深圳）","sort_order":2},{"forum_id":4190,"role":"guest","name":"王希廷","affiliation":"中国人民大学","sort_order":3},{"forum_id":4190,"role":"guest","name":"徐汉辉","affiliation":"复旦大学","sort_order":4}],"host":"","image_url":"/api/forums/4190/cover"},{"id":4191,"title":"Third Forum on Multi-Agent Collaboration and Game Theory: Frontiers of Embodied Multi-Agent Systems","topic":"AI Agents","kind":"专题论坛","chair":"魏巍","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-513","description":"Building on the previous two editions' focus on multi-agent collaboration and game theory, this forum extends its research perspective into the physical world, exploring collective perception, collaborative interaction, and adversarial game play among multiple agents in real-world scenarios. The forum guides panelists to deliberate on core issues such as \"how to achieve robust collaborative perception in real physical environments,\" \"how embodied agents can carry out efficient physical collaboration,\" and \"how world models can enhance multi-agent collaboration and game-playing capabilities.\" The rationale for recommending this topic is that, with the rapid development of fields such as humanoid robots, autonomous driving, and drones, multi-agent embodied collaboration has become an important path to breaking through the capability bottlenecks of single machines. This direction deeply integrates technologies such as large models, multimodal perception, and reinforcement learning, and is at the intersection of theoretical innovation and industrial deployment. The forum aims to bring together academia, industry, and research forces to explore technological breakthroughs and application paradigms for embodied multi-agent systems, which is of great significance for advancing artificial intelligence from the virtual world into the physical world.","is_demo":0,"cover_url":"/api/forums/4191/cover?size=thumb","cover_alt":"Third Forum on Multi-Agent Collaboration and Game Theory: Frontiers of Embodied Multi-Agent Systems","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4191","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening\",\"魏巍\"],[\"Welcome Remarks\",\"梁吉业\"],[\"Embodied Intelligence and General Behavioral Game Evolution of Multi-Agent Systems\",\"孙富春\"],[\"Multi-Task-Oriented Embodied Multi-Agent Decision-Making\",\"赵冬斌\"],[\"Autonomous UAV Perception, Decision-Making, and Control in Complex Dynamic Environments\",\"秦家虎\"],[\"Aerial Embodied Intelligence and Multi-Arm Cooperative Aerial Manipulation\",\"吴奇\"],[\"Key Technologies and Applications of Intelligent Perception in Low-Altitude Environments\",\"朱鹏飞\"]]","code":"CNCC2026-4191","forum_type":"专题论坛","format":null,"theme_code":"Agent智能体","theme_name":"Agent智能体","venue":"西博城5号馆","room":"西博城5号馆-513","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":7,"forum_id":"CNCC2026-4191","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第三届多智能体协同与博弈论坛--具身多智能体系统前沿","original_topic":"Agent智能体","original_description":"本届论坛在延续前两届聚焦多智能体协作与博弈的基础上，将研究视角延伸至物理世界，探讨真实场景下多智能体的群体感知、协作交互与对抗博弈。论坛引导嘉宾围绕“真实物理环境下如何实现鲁棒协同感知”“具身智能体如何开展高效物理协作”“世界模型如何提升多智能体协同与博弈能力”等核心议题展开思辨。推荐本选题的理由在于，随着人形机器人、自动驾驶及无人机等领域的快速发展，多智能体具身协同已成为突破单机能力瓶颈的重要路径。该方向深度融合大模型、多模态感知与强化学习等技术，正处于理论创新与产业落地的交汇期。论坛旨在汇聚产学研力量，探索具身多智能体系统的技术突破与应用范式，对推动人工智能从虚拟走向物理世界具有重要意义。","people":[{"forum_id":4191,"role":"chair","name":"魏巍","affiliation":"山西大学","sort_order":0},{"forum_id":4191,"role":"guest","name":"梁吉业","affiliation":"山西大学","sort_order":0},{"forum_id":4191,"role":"guest","name":"孙富春","affiliation":"清华大学","sort_order":1},{"forum_id":4191,"role":"guest","name":"赵冬斌","affiliation":"中国科学院自动化研究所","sort_order":2},{"forum_id":4191,"role":"guest","name":"秦家虎","affiliation":"中国科学技术大学","sort_order":3},{"forum_id":4191,"role":"guest","name":"吴奇","affiliation":"上海交通大学","sort_order":4},{"forum_id":4191,"role":"guest","name":"朱鹏飞","affiliation":"东南大学","sort_order":5},{"forum_id":4191,"role":"guest","name":"袁雷","affiliation":"南京大学","sort_order":6}],"host":"","image_url":"/api/forums/4191/cover"},{"id":4194,"title":"Forum on Space-Based Computing and Intelligent Applications","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"韩银和","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-406","description":"This forum focuses on key technologies of in-orbit intelligent computing and highly reliable computing, and conducts discussions around directions such as aerospace computing systems, space-grade highly reliable devices, on-board intelligent computing, heterogeneous computing, and satellite-ground collaboration. The forum will focus on the construction of high-computing-power, highly reliable computing platforms in space environments, on-board intelligent processing and hardware-software co-design, and the application challenges of key technologies in aerospace missions such as remote sensing, communication, navigation, and deep space exploration. By sharing frontier research results and engineering practices, it promotes academic exchange and industry-university-research cooperation, drives key technological breakthroughs and achievement transformation, and provides technical support for the construction of intelligent constellations and the intelligent development of aerospace information systems.","is_demo":0,"cover_url":"/api/forums/4194/cover?size=thumb","cover_alt":"Forum on Space-Based Computing and Intelligent Applications","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4194","source_checked_at":"2026-10-30","agenda_json":"[[\"Self-Organizing Broadband Wireless Access Technology\",\"李建东\"],[\"Space-Borne Transport Networks and Computing-Network Collaboration Trends\",\"陈山枝\"],[\"Space-Based Intelligent Computing Constellation Architecture and Its Key Technologies\",\"汪自军\"],[\"Exploration of Space-Based Large Models and Agents\",\"赵益晟\"]]","code":"CNCC2026-4194","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-406","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4194","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"天基计算与智能应用论坛","original_topic":"新型计算与网络架构","original_description":"本论坛聚焦在轨智能计算与高可靠计算关键技术，围绕空天计算体系、航天高可靠器件、星上智能计算、异构计算与星地协同等方向展开研讨。论坛将重点探讨空间环境下高算力、高可靠计算平台构建，星上智能处理与软硬件协同设计，以及关键技术在遥感、通信、导航、深空探测等航天任务中的应用挑战。通过分享前沿研究成果和工程实践，促进学术交流与产学研合作，推动关键技术突破与成果转化，为智能星座建设和空天信息系统智能化发展提供技术支撑。","people":[{"forum_id":4194,"role":"chair","name":"韩银和","affiliation":"中国科学院计算技术研究所","sort_order":0},{"forum_id":4194,"role":"guest","name":"李建东","affiliation":"西安电子科技大学","sort_order":0},{"forum_id":4194,"role":"guest","name":"陈山枝","affiliation":"中国信息通信科技集团有限公司","sort_order":1},{"forum_id":4194,"role":"guest","name":"汪自军","affiliation":"中国商业卫星上海航天空间技术有限公司","sort_order":2},{"forum_id":4194,"role":"guest","name":"赵益晟","affiliation":"国科大杭州高等研究院","sort_order":3},{"forum_id":4194,"role":"guest","name":"Tony QUEK","affiliation":"新加坡科技设计大学","sort_order":4},{"forum_id":4194,"role":"guest","name":"华更新","affiliation":"北京控制工程研究所","sort_order":5},{"forum_id":4194,"role":"guest","name":"刘垚圻","affiliation":"中国科学院计算技术研究所","sort_order":6},{"forum_id":4194,"role":"guest","name":"姜兴龙","affiliation":"中科天算信息科技有限公司","sort_order":7},{"forum_id":4194,"role":"guest","name":"孟繁涛","affiliation":"中科天算信息科技有限公司","sort_order":8}],"host":"","image_url":"/api/forums/4194/cover"},{"id":4195,"title":"From Geometric Computing to Spatial Intelligence: Perception, Generation and Interaction of 3D Content","topic":"Vision, Multimodality and World Models","kind":"专题论坛","chair":"高林","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-413","description":"With the rapid development of generative artificial intelligence, 3D vision and embodied intelligence, how to enable machines to understand, reconstruct and create the real 3D world has become an important frontier at the intersection of computer graphics and artificial intelligence. This forum will hold discussions around directions such as 3D geometric representation and computing, perception and modeling of complex scenes, 3D content generation and visualization, human-computer interaction and embodied intelligence, connecting foundational theories such as computational conformal geometry and topology, data-driven 3D perception and generation methods, and spatial intelligence technologies for the digital world and robotic applications. The forum aims to explore the development trends, core challenges and future opportunities of 3D intelligence from foundational theories and key algorithms to practical applications.","is_demo":0,"cover_url":"/api/forums/4195/cover?size=thumb","cover_alt":"From Geometric Computing to Spatial Intelligence: Perception, Generation and Interaction of 3D Content","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4195","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Remarks\",\"高林\"],[\"Physically Consistent 3D Spatial Generation and Simulation Learning\",\"鲍虎军\"],[\"Graphics Simulation and World Models\",\"陈宝权\"],[\"A Large-Scale Scientific Facility for Life Intelligence Detection from a 3D Perspective\",\"罗笑南\"],[\"Multi-Step Cognitive Reasoning over Spatial Scenes Based on Chain-of-Thought\",\"马利庄\"],[\"From Pixels to Compute: The Co-Evolution of GPUs and Graphics\",\"周昆\"]]","code":"CNCC2026-4195","forum_type":"专题论坛","format":null,"theme_code":"视觉、多模态与世界模型","theme_name":"视觉、多模态与世界模型","venue":"西博城4号馆","room":"西博城4号馆-413","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4195","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"从几何计算到空间智能：三维内容的感知、生成与交互","original_topic":"视觉、多模态与世界模型","original_description":"随着生成式人工智能、三维视觉和具身智能快速发展，如何让机器理解、重建并创造真实三维世界，已成为计算机图形学与人工智能交叉领域的重要前沿。本论坛将围绕三维几何表征与计算、复杂场景感知与建模、三维内容生成与可视化、人机交互及具身智能等方向展开研讨，贯通计算共形几何与拓扑等基础理论、数据驱动的三维感知与生成方法，以及面向数字世界和机器人应用的空间智能技术。论坛旨在探讨三维智能从基础理论、关键算法到实际应用的发展趋势、核心挑战与未来机遇。","people":[{"forum_id":4195,"role":"chair","name":"高林","affiliation":"中国科学院计算技术研究所","sort_order":0},{"forum_id":4195,"role":"guest","name":"鲍虎军","affiliation":"浙江大学CAD&CG国家重点实验室","sort_order":0},{"forum_id":4195,"role":"guest","name":"陈宝权","affiliation":"北京大学","sort_order":1},{"forum_id":4195,"role":"guest","name":"罗笑南","affiliation":"桂林电子科技大学","sort_order":2},{"forum_id":4195,"role":"guest","name":"马利庄","affiliation":"上海交通大学","sort_order":3},{"forum_id":4195,"role":"guest","name":"周昆","affiliation":"浙江大学","sort_order":4}],"host":"","image_url":"/api/forums/4195/cover"},{"id":4197,"title":"The Second Humanities Intelligence Frontier Forum—Humanities Intelligence and Cultural Big Data Forum","topic":"General","kind":"专题论坛","chair":"吴亚东","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-108邛海","description":"With the rapid development of artificial intelligence technology, the integration of artificial intelligence and cultural big data is becoming an important engine for driving interdisciplinary innovation. This forum takes \"Humanities Intelligence and Cultural Big Data\" as its theme, focusing on the knowledge organization methodology of cultural big data and its application in regional cultural research.\nThe forum will explore frontier methods of humanities intelligence, focusing on how emerging technologies can serve the deep understanding and revitalized utilization of cultural resources; at the same time, it will exchange views on the organization models and standards construction of cultural big data, and explore data-driven research paths. The forum especially takes Sichuan's poetry and wine culture as an entry point to discuss how to build thematic datasets that combine cultural depth and data quality, promoting the digital-intelligent preservation and intelligent dissemination of local distinctive culture.\nThe forum aims to build an interdisciplinary dialogue platform, promote the research of artificial intelligence and cultural big data from methodological exploration to implementable practical solutions, and provide academic support for the inheritance and innovation of fine traditional Chinese culture.","is_demo":0,"cover_url":"/api/forums/4197/cover?size=thumb","cover_alt":"The Second Humanities Intelligence Frontier Forum—Humanities Intelligence and Cultural Big Data Forum","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4197","source_checked_at":"2026-10-30","agenda_json":"[[\"Building the National Cultural Big Data System: Top-Level Design and Practical Exploration\",\"高书生\"],[\"Data-Driven Visual Analysis of Historical and Archaeological Materials\",\"袁晓如\"],[\"From Interpersonal Collaboration to Human-AI Collaboration: Shifts in Historical Research Approaches in the AI Era\",\"梁晨\"],[\"From \\\"Fragments\\\" to \\\"Glyphs\\\": Intelligent Restoration, Analysis, and Generation of Cultural Heritage Images\",\"彭舰\"],[\"Practices and University-Industry Collaboration Pathways for Humanities Intelligence-Empowered High-Quality Cultural Dataset Construction: A Core Case Study of the Poetry and Wine Culture Thematic Database\",\"郎毅\"]]","code":"CNCC2026-4197","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"天府国际会议中心","room":"天府国际会议中心-108邛海","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4197","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第二届人文智能前沿论坛——人文智能与文化大数据论坛","original_topic":"综合","original_description":"随着人工智能技术的快速发展，人工智能与文化大数据的融合正成为推动跨学科创新的重要引擎。本届论坛以\"人文智能与文化大数据\"为主题，聚焦文化大数据的知识组织方法论及其在区域文化研究中的应用。\n论坛将围绕人文智能的前沿方法展开探讨，关注新兴技术如何服务于文化资源的深层理解与活化利用；同时就文化大数据的组织模式与标准建设进行交流，探索数据驱动的研究路径。论坛特别以四川诗酒文化为切入点，探讨如何构建兼具文化深度与数据品质的专题数据集，推动地方特色文化的数智化保存与智能化传播。\n论坛旨在搭建跨学科对话平台，推动人工智能与文化大数据研究从方法探索走向可落地的实践方案，为中华优秀传统文化的传承创新提供学术支撑。","people":[{"forum_id":4197,"role":"chair","name":"吴亚东","affiliation":"四川轻化工大学","sort_order":0},{"forum_id":4197,"role":"guest","name":"高书生","affiliation":"中央宣传部","sort_order":0},{"forum_id":4197,"role":"guest","name":"袁晓如","affiliation":"北京大学智能学院","sort_order":1},{"forum_id":4197,"role":"guest","name":"梁晨","affiliation":"南京大学","sort_order":2},{"forum_id":4197,"role":"guest","name":"彭舰","affiliation":"四川大学","sort_order":3},{"forum_id":4197,"role":"guest","name":"郎毅","affiliation":"四川省文化大数据有限责任公司","sort_order":4},{"forum_id":4197,"role":"guest","name":"高原","affiliation":"西北大学","sort_order":5},{"forum_id":4197,"role":"guest","name":"拥措","affiliation":"西藏大学","sort_order":6},{"forum_id":4197,"role":"guest","name":"杨晓飞","affiliation":"中国科学院上海高等研究院","sort_order":7},{"forum_id":4197,"role":"guest","name":"左家莉","affiliation":"江西师范大学","sort_order":8},{"forum_id":4197,"role":"guest","name":"战蓓蓓","affiliation":"湖南大学","sort_order":9},{"forum_id":4197,"role":"guest","name":"陈志威","affiliation":"南昌大学","sort_order":10}],"host":"","image_url":"/api/forums/4197/cover"},{"id":4202,"title":"How AI Agents Are \"Tearing Apart\" Traditional Protection Architectures: Paradigm Reconstruction and Technical Responses for Multi-Domain Converged Security Systems","topic":"AI Safety and Governance","kind":"专题论坛","chair":"饶志宏、白杨","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-505","description":"As AI Agents move from assisting decision-making to autonomous execution, cyber offense and defense are accelerating toward cloud-edge-device collaboration, the interweaving of physical, electromagnetic, and cyber cognition, and multi-scenario applications such as the industrial Internet, satellite Internet, low-altitude economy, and embodied intelligence. While driving a leap in cybersecurity capabilities, this also brings risks of tearing, such as model trustworthiness, AI Agent collaboration, and loss of permission control. Based on the development needs of the cybersecurity industry in the national strategic hinterland, this forum focuses on AI Agent-driven multi-domain converged cyber offense and defense, conducting in-depth deliberation around capability reconstruction, risk evolution, and security system reshaping, and jointly exploring a new cybersecurity paradigm in the era of AI Agents.\n\nTopic 1: How are AI Agents reconstructing cyber offense and defense? A capability leap or an efficiency revolution?\nTopic 2: Will AI Agents tear apart existing cyber offense and defense security systems? Do the risks come from models, AI Agents, or the collaborative ecosystem?\nTopic 3: How can we reshape a trustworthy and controllable multi-domain converged cyber offense and defense security system driven by AI Agents?","is_demo":0,"cover_url":"/api/forums/4202/cover?size=thumb","cover_alt":"How AI Agents Are \"Tearing Apart\" Traditional Protection Architectures: Paradigm Reconstruction and Technical Responses for Multi-Domain Converged Security Systems","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4202","source_checked_at":"2026-10-30","agenda_json":"[[\"Inertia, South Wall, Lane Change: A New Starting Point for Cybersecurity\",\"王宝生\"],[\"Intelligent Vulnerability Discovery for Network Device Firmware\",\"张超\"],[\"Exploring Cyber Threat Mechanisms and Building Behavior Gene Datasets\",\"陈兴蜀\"]]","code":"CNCC2026-4202","forum_type":"专题论坛","format":null,"theme_code":"AI安全与治理","theme_name":"AI安全与治理","venue":"西博城5号馆","room":"西博城5号馆-505","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":3,"forum_id":"CNCC2026-4202","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI智能体如何“撕裂”传统防护架构？多域融合安全体系范式重构与技术应对","original_topic":"AI安全与治理","original_description":"随着AI智能体从辅助决策迈向自主执行，网络攻防正加速向云边端协同、物理电磁网络认知交织以及工业互联网、卫星互联网、低空经济、具身智能等多场景应用演进，推动网安能力跃迁同时，也带来模型可信、智能体协同、权限失控等风险撕裂。本论坛立足国家战略腹地网安产业发展需求，聚焦AI智能体驱动的多域融合网络攻防，围绕能力重构、风险演化与安全体系重塑开展深度思辨，共探智能体时代网络安全新范式。\n\n议题一：AI智能体正在如何重构网络攻防？能力跃迁还是效率革命？\n议题二：AI智能体会撕裂现有网络攻防安全体系吗？风险来自模型、智能体还是协同生态？\n议题三：如何重塑可信可控AI智能体驱动的多域融合网络攻防安全体系？","people":[{"forum_id":4202,"role":"chair","name":"饶志宏","affiliation":"中国电子科技集团有限公司","sort_order":0},{"forum_id":4202,"role":"guest","name":"吴世忠","affiliation":"国家保密局","sort_order":0},{"forum_id":4202,"role":"co_chair","name":"白杨","affiliation":"成都信息工程大学","sort_order":1},{"forum_id":4202,"role":"guest","name":"张宝东","affiliation":"战略支援部队某技术研究中心","sort_order":1},{"forum_id":4202,"role":"guest","name":"季向阳","affiliation":"清华大学","sort_order":2},{"forum_id":4202,"role":"guest","name":"王宝生","affiliation":"国防科技大学","sort_order":3},{"forum_id":4202,"role":"guest","name":"张超","affiliation":"清华大学","sort_order":4},{"forum_id":4202,"role":"guest","name":"陈兴蜀","affiliation":"四川大学信息管理中心","sort_order":5},{"forum_id":4202,"role":"guest","name":"方乐缘","affiliation":"湖南大学","sort_order":6}],"host":"","image_url":"/api/forums/4202/cover"},{"id":4204,"title":"Wireless Short-Range Networks in the AI Era: Evolution or Reconstruction?","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"迟国轩、徐方鑫","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-105A都江堰A","description":"In the past, AI was mainly used as an optimization tool for wireless networks, for parameter configuration, resource scheduling, and operations management, i.e., \"AI for Network\". As large models and AI agents gradually become new objects of wireless network connectivity and services, \"Network for AI\" begins to bring new requirements such as task collaboration, context transfer, and distributed inference. Can existing wireless networks meet these requirements through protocol enhancements and architectural evolution, or is it necessary to redesign connection mechanisms, protocol architectures, and hardware-software systems for AI? This forum will invite experts in wireless networks, artificial intelligence, chip systems, and industry standards to exchange views around \"incremental evolution or systemic reconstruction\", compare the gains and losses of the two paths in terms of performance benefits, compatibility costs, reliability and security, and standardization, clarify the scenarios and boundaries that require reconstruction, and form a staged consensus on near-term technical breakthroughs and standardization layout.","is_demo":0,"cover_url":"/api/forums/4204/cover?size=thumb","cover_alt":"Wireless Short-Range Networks in the AI Era: Evolution or Reconstruction?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4204","source_checked_at":"2026-10-30","agenda_json":"[[\"Agent Interconnection and Its Governance: Open and Win-Win, Driving Innovation\",\"李珂\"],[\"Agentic AI Empowering Short-Range Communication Testing and Industry-Academia-Research Innovation\",\"曲雅江\"]]","code":"CNCC2026-4204","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"天府国际会议中心","room":"天府国际会议中心-105A都江堰A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":2,"forum_id":"CNCC2026-4204","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI时代的无线短距网络：演进还是重构？","original_topic":"新型计算与网络架构","original_description":"过去，AI主要作为无线网络的优化工具，用于参数配置、资源调度和运维管理，即“AI for Network”。随着大模型与智能体逐渐成为无线网络连接和服务的新对象，“Network for AI”开始带来任务协同、上下文传递与分布式推理等新需求。现有无线网络通过协议增强和架构演进能否满足这些需求，还是需要面向AI重新设计连接机制、协议架构与软硬件体系？本论坛将邀请无线网络、人工智能、芯片系统与产业标准领域专家，围绕“渐进演进还是体系重构”展开观点交锋，比较两条路径在性能收益、兼容成本、可靠安全和标准化方面的得失，厘清需要重构的场景与边界，形成近期技术突破和标准布局的阶段性共识。","people":[{"forum_id":4204,"role":"chair","name":"迟国轩","affiliation":"北京邮电大学","sort_order":0},{"forum_id":4204,"role":"guest","name":"欧阳武","affiliation":"中国通信学会","sort_order":0},{"forum_id":4204,"role":"co_chair","name":"徐方鑫","affiliation":"云程半导体","sort_order":1},{"forum_id":4204,"role":"guest","name":"王田","affiliation":"北京师范大学","sort_order":1},{"forum_id":4204,"role":"guest","name":"徐展聪","affiliation":"MTEL 电信有限公司","sort_order":2},{"forum_id":4204,"role":"guest","name":"李珂","affiliation":"北京邮电大学","sort_order":3},{"forum_id":4204,"role":"guest","name":"曲雅江","affiliation":"灿芯技术（深圳）有限公司","sort_order":4},{"forum_id":4204,"role":"guest","name":"华忠","affiliation":"是德科技","sort_order":5},{"forum_id":4204,"role":"guest","name":"朱天清","affiliation":"澳门城市大学","sort_order":6},{"forum_id":4204,"role":"guest","name":"向乔","affiliation":"厦门大学","sort_order":7},{"forum_id":4204,"role":"guest","name":"张磊","affiliation":"河南大学","sort_order":8},{"forum_id":4204,"role":"guest","name":"刘应波","affiliation":"云南财经大学","sort_order":9},{"forum_id":4204,"role":"guest","name":"高雅玙","affiliation":"华中科技大学","sort_order":10},{"forum_id":4204,"role":"guest","name":"杨旭东","affiliation":"西安工程大学","sort_order":11},{"forum_id":4204,"role":"guest","name":"胡钊鸣","affiliation":"太原理工大学","sort_order":12}],"host":"","image_url":"/api/forums/4204/cover"},{"id":4205,"title":"From Sci-Fi to Future Computing: Technological Imagination and Innovation Boundaries in the AI Era","topic":"General","kind":"专题论坛","chair":"王元卓","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城2号馆-206","description":"Science fiction is a thought experiment ground for technology, providing computing science with imaginative frameworks that break conventional boundaries; meanwhile, frontier advances such as quantum computing and brain-inspired intelligence are accelerating sci-fi scenarios into reality. This forum invites computing scientists and science fiction writers to engage in cross-disciplinary dialogue, focusing on three directions: How can science fiction narratives inspire real research paths? How do technological breakthroughs reshape the logic of science fiction creation? How can the power of science fiction be leveraged to promote the \"breaking out of the circle\" dissemination of computing science popularization?\nThe forum will raise a series of speculative questions: Under what conditions can science fiction imagination be transformed into research problems? When AI begins to create, will human imagination be replaced? How can a two-way empowerment mechanism be established between scientists and writers?\nThis forum aims to enhance CCF's overall awareness of science popularization work and comprehensive science popularization innovation capability through CNCC, while also providing CCF branches with a replicable paradigm for science popularization activities. Cross-disciplinary dialogue stimulates innovative inspiration, and the collision of ideas empowers science popularization work.","is_demo":0,"cover_url":"/api/forums/4205/cover?size=thumb","cover_alt":"From Sci-Fi to Future Computing: Technological Imagination and Innovation Boundaries in the AI Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4205","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4205","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城2号馆","room":"西博城2号馆-206","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":0,"forum_id":"CNCC2026-4205","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"从科幻到未来计算:AI时代的技术想象与创新边界","original_topic":"综合","original_description":"科幻是技术的思想实验场，为计算科学提供突破常规的想象框架；而量子计算、类脑智能等前沿进展，又让科幻设定加速走进现实。本论坛邀请计算领域科学家与科幻作家跨界对话，聚焦三大方向：科幻叙事如何启发真实科研路径？技术突破如何重塑科幻创作逻辑？如何借助科幻力量推动计算科普的“破圈”传播？\n论坛将提出系列思辨问题：科幻想象在什么条件下能转化为科研问题？当AI开始创作，人类想象力是否会被替代？科学家与作家之间如何建立双向赋能机制？\n本论坛旨在通过CNCC提升CCF整体对科普工作的认知和全面科普创新能力，同时为CCF分支机构提供可借鉴的科普活动范式。跨界对话激发创新灵感，思想碰撞赋能科普工作。","people":[{"forum_id":4205,"role":"chair","name":"王元卓","affiliation":"中国科学院计算技术研究所","sort_order":0},{"forum_id":4205,"role":"guest","name":"姜文婷","affiliation":"破晓云依科技公司","sort_order":0},{"forum_id":4205,"role":"guest","name":"陈晨","affiliation":"北京科学中心","sort_order":1},{"forum_id":4205,"role":"guest","name":"严曦","affiliation":"科幻作家","sort_order":2},{"forum_id":4205,"role":"guest","name":"张文强","affiliation":"复旦大学","sort_order":3},{"forum_id":4205,"role":"guest","name":"杨国梁","affiliation":"四川科幻世界杂志社","sort_order":4},{"forum_id":4205,"role":"guest","name":"王诺诺","affiliation":"《科幻世界》杂志社","sort_order":5}],"host":"","image_url":"/api/forums/4205/cover"},{"id":4207,"title":"From Tokens to Value: Building and Practicing Sustainable AI Systems and Products","topic":"Unclassified","kind":"专题论坛","chair":"段亦涛","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-516","description":"This forum focuses on the key transformation path \"from tokens to value,\" exploring how to build sustainable artificial intelligence systems and products in the era of large models. The topics will revolve around token cost and efficiency optimization, the capability evolution of foundation large models, the design and optimization of agent systems, and the construction of agent products and business models. The forum will combine product cases such as token platform cost reduction and efficiency improvement practices, academic research agents, and Lobster general agents, and invite technical experts from top foundation large model companies to share their experience in implementing foundation models and agents. Through multidimensional discussion of technical architecture, product experience, and business value, the forum will jointly answer: how to transform model invocation costs into real user value, how to move agents from capability demonstrations to stable and usable products, and how to establish a balance among efficiency, effectiveness, and sustainable growth.","is_demo":0,"cover_url":"/api/forums/4207/cover?size=thumb","cover_alt":"From Tokens to Value: Building and Practicing Sustainable AI Systems and Products","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4207","source_checked_at":"2026-10-30","agenda_json":"[[\"MiniMax - AI Native Organizational Practice: Everyone Is an Agent Designer\",\"葛明放\"],[\"Agents and Domain-Specific Languages\",\"汪源\"],[\"From Tokens to Trusted Value Construction: Governance and Practice for Trustworthy AI\",\"韩蒙\"],[\"Best Practices for On-Device Model Applications in the AMD Hardware and Software Ecosystem\",\"杨广远\"],[\"Reflections and Practice on Agent Products Under the AI-Native Strategy\",\"段亦涛\"]]","code":"CNCC2026-4207","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-516","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4207","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"从Token到价值：可持续人工智能系统和产品的构建与实践","original_topic":"未分类","original_description":"本论坛聚焦“从Token到价值”的关键转化路径，探讨大模型时代如何构建可持续的人工智能系统与产品。议题将围绕Token成本与效率优化、基座大模型能力演进、智能体系统设计与优化、智能体产品构建及商业模式展开。论坛将结合Token平台降本增效实践、学术研究智能体、龙虾通用智能体等产品案例，并邀请顶部基座大模型公司技术专家分享基座模型与智能体落地经验。通过技术架构、产品体验和商业价值的多维讨论，论坛将共同回答：如何让模型调用成本转化为真实用户价值，如何让智能体从能力展示走向稳定可用的产品，如何在效率、效果与可持续增长之间建立平衡。","people":[{"forum_id":4207,"role":"chair","name":"段亦涛","affiliation":"网易有道信息技术（北京）有限公司","sort_order":0},{"forum_id":4207,"role":"guest","name":"葛明放","affiliation":"MiniMax（上海稀宇科技有限公司）","sort_order":0},{"forum_id":4207,"role":"guest","name":"汪源","affiliation":"杭州久痕科技有限公司","sort_order":1},{"forum_id":4207,"role":"guest","name":"韩蒙","affiliation":"浙江大学","sort_order":2},{"forum_id":4207,"role":"guest","name":"杨广远","affiliation":"超威半导体 （AMD）","sort_order":3},{"forum_id":4207,"role":"guest","name":"张广勇","affiliation":"网易有道信息技术（北京）有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4207/cover"},{"id":4210,"title":"AI-Driven Vulnerability Discovery: From Tool Assistance to AI Agent Practice","topic":"AI Safety and Governance","kind":"专题论坛","chair":"张玉清","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-507","description":"As the reasoning, planning, tool-calling, and autonomous decision-making capabilities of large language models continue to improve, AI Agents are driving vulnerability discovery toward intelligence and automation. The core topics include: first, outlining the capability foundations and system mechanisms of AI Agent-empowered vulnerability discovery; second, sharing representative systems and frontier advances of AI Agents in key vulnerability discovery tasks, and discussing their integrated application with traditional vulnerability discovery techniques; third, discussing the engineering practice, capability boundaries, and development trends of AI Agents in real software systems.\nRecommended reason: The complexity of software systems continues to increase, and traditional vulnerability discovery methods face efficiency and scale challenges. AI Agents are reconstructing vulnerability discovery and security analysis processes. This forum will bring together the latest achievements in related fields, promote exchange and cooperation between academia and industry, and drive the innovation and application of AI Agent-empowered cybersecurity technologies.","is_demo":0,"cover_url":"/api/forums/4210/cover?size=thumb","cover_alt":"AI-Driven Vulnerability Discovery: From Tool Assistance to AI Agent Practice","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4210","source_checked_at":"2026-10-30","agenda_json":"[[\"Reflections on Vulnerability Discovery Technology in the Era of AI Agents\",\"陈恺\"],[\"Large Model-Driven Network Penetration Task Orchestration and Error Correction\",\"纪守领\"],[\"Large Model-Driven Application Security Analysis: Practical Exploration from Vulnerability Detection to Exploitability Verification\",\"刘奇旭\"],[\"Practices in Defect Detection for AI Agent Systems\",\"梁彬\"],[\"Reflections and Practices on Fuzzing Technology in the AI Era\",\"于冬松\"]]","code":"CNCC2026-4210","forum_type":"专题论坛","format":null,"theme_code":"AI安全与治理","theme_name":"AI安全与治理","venue":"西博城5号馆","room":"西博城5号馆-507","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4210","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI驱动漏洞发现：从工具辅助到智能体实战","original_topic":"AI安全与治理","original_description":"随着大语言模型推理、规划、工具调用和自主决策能力不断提升，AI Agent正推动漏洞挖掘向智能化、自动化发展。核心议题包括：一是梳理AI Agent赋能漏洞挖掘的能力基础与系统机制；二是分享AI Agent在漏洞挖掘关键任务中的代表性系统与前沿进展，探讨其与传统漏洞挖掘技术的融合应用；三是探讨AI Agent在真实软件系统中的工程实践、能力边界及发展趋势。\n推荐理由：软件系统复杂度持续提升，传统漏洞挖掘方法面临效率与规模挑战。AI Agent正在重构漏洞挖掘与安全分析流程。本论坛将汇聚相关领域最新成果，促进学术界与产业界交流合作，推动AI Agent赋能网络安全技术创新与应用。","people":[{"forum_id":4210,"role":"guest","name":"陈恺","affiliation":"中国科学院信息工程研究所","sort_order":0},{"forum_id":4210,"role":"chair","name":"张玉清","affiliation":"中国科学院大学","sort_order":0},{"forum_id":4210,"role":"guest","name":"纪守领","affiliation":"浙江大学","sort_order":1},{"forum_id":4210,"role":"guest","name":"刘奇旭","affiliation":"中国科学院信息工程研究所","sort_order":2},{"forum_id":4210,"role":"guest","name":"梁彬","affiliation":"中国人民大学","sort_order":3},{"forum_id":4210,"role":"guest","name":"于冬松","affiliation":"中关村实验室","sort_order":4},{"forum_id":4210,"role":"guest","name":"李振源","affiliation":"浙江大学","sort_order":5}],"host":"","image_url":"/api/forums/4210/cover"},{"id":4211,"title":"AI-Powered Smart Medical and Elderly Care: Examining the \"Dao, Techniques, and Trends\" of Digital Health","topic":"Unclassified","kind":"专题论坛","chair":"陈立明","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-514","description":"\"Care for the elderly and medical treatment for the sick\" has become a major social need for Healthy China. Under limited medical and elderly care resources, AI-powered smart medical and elderly care provides a technological approach to addressing this challenge. Although medical and elderly care services are shifting from \"passive diagnosis and treatment, separation of medical care and elderly care\" to an integrated prevention-treatment-care model of \"prevention as the priority, precision diagnosis and treatment, and integration of medical care and elderly care,\" problems remain, such as fragmented technologies, shortages of medical data, and difficulties in implementing application scenarios. Faced with rapidly developing science and technology and continuously evolving user needs, realizing AI-powered smart medical and elderly care requires new ideas and new methods.\nThis forum conducts a multidimensional and systematic discussion around the \"Dao, techniques, and trends\" of AI-powered smart medical and elderly care. \"Dao\" refers to the core concepts and paradigms; \"techniques\" refers to methods and technologies; \"trends\" refers to the current situation and trends. The forum gathers top experts in related fields to discuss people-centered and scenario-driven smart health management that builds an integrated prevention-treatment-care system both inside and outside hospitals; key technologies such as multimodal intelligent perception and behavior understanding, intelligent spatial modeling and reasoning, medical large models, autonomous decision-making by agents, human-AI collaboration, and embodied services; and, combined with the latest research practices in smart homes, chronic disease management, and wireless sensing, to explore new paradigms, frontier technologies, and innovative practices for AI-powered smart medical and elderly care, pointing the way toward a scalable and personalized future smart medical and elderly care system.","is_demo":0,"cover_url":"/api/forums/4211/cover?size=thumb","cover_alt":"AI-Powered Smart Medical and Elderly Care: Examining the \"Dao, Techniques, and Trends\" of Digital Health","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4211","source_checked_at":"2026-10-30","agenda_json":"[[\"The \\\"Dao, Techniques, and Trends\\\" of AI-Empowered Smart Healthcare and Elderly Care\",\"陈立明\"],[\"AI-Based Analysis of Biomedical Big Data and Its Application to Early-Stage Autism Grading Screening\",\"潘毅\"],[\"Artificial Intelligence in the Diagnosis and Treatment of Mental Disorders\",\"胡斌\"],[\"From Brain Mechanism Analysis to Cognitive Decoding\",\"张道强\"],[\"Millimeter-Wave/Terahertz Fine-Grained Sensing for Non-Contact Health Monitoring\",\"周安福\"]]","code":"CNCC2026-4211","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-514","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4211","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI智慧医养：把脉数字健康之“道、术与势”","original_topic":"未分类","original_description":"“老有所养、病有所医”成为健康中国的重大社会需求，在有限的医疗和养老资源下，AI智慧医养为应对这一挑战提供了技术解决途径。虽然医疗与养老服务正由“被动诊疗、医养分离”向“以防为主、精准诊疗、医养结合”的防—治—养一体化模式转变。但仍存在技术碎片化、医疗数据短缺、应用场景落地难等问题。面对飞速发展的科学技术及不断演绎的用户需求，实现AI智慧医养需要新思路和新方法。\n本论坛围绕AI智慧医养的“道、术与势”开展多维度系统探讨。“道”是核心理念与范式；“术“是方法与技术；“势”是时势与趋势。论坛集聚相关领域顶尖专家，围绕以人为中心和场景驱动、构建院内与院外防—治—养一体化的智慧健康管理；多模态智能感知与行为理解、智能空间建模与推理、医学大模型、智能体自主决策、人智协同及具身服务等关键技术；并结合智能家居、慢病管理和无线感知等最新科研实践，探讨AI智慧医养新范式、前沿技术和创新实践，为实现规模化、个性化的未来智慧医养体系指明方向。","people":[{"forum_id":4211,"role":"chair","name":"陈立明","affiliation":"大连理工大学","sort_order":0},{"forum_id":4211,"role":"guest","name":"潘毅","affiliation":"深圳理工大学","sort_order":0},{"forum_id":4211,"role":"guest","name":"胡斌","affiliation":"北京理工大学","sort_order":1},{"forum_id":4211,"role":"guest","name":"张道强","affiliation":"南京航空航天大学","sort_order":2},{"forum_id":4211,"role":"guest","name":"周安福","affiliation":"北京邮电大学计算机学院","sort_order":3},{"forum_id":4211,"role":"guest","name":"方玲玲","affiliation":"辽宁师范大学","sort_order":4}],"host":"","image_url":"/api/forums/4211/cover"},{"id":4212,"title":"How Can AI Enter the Real World? From Transportation Large Models and Agents to Autonomous Systems","topic":"AI for Industry","kind":"专题论坛","chair":"陈婧、曹原周汉","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-207A天府厅A","description":"Artificial intelligence is currently moving from content generation and knowledge question answering in digital space toward environmental perception, task planning, interactive decision-making, and autonomous operation in the real world. Transportation systems feature multi-source heterogeneous data, complex spatiotemporal correlations, dynamic open environments, strong real-time requirements, and safety constraints, making them an important scenario for testing the real capabilities of large models and agents. This forum focuses on issues such as transportation large models, agent decision-making, world models, simulation evaluation, trustworthy autonomy, and device-edge-cloud collaboration, and explores how artificial intelligence can evolve from single-point algorithms and auxiliary tools into autonomous intelligent systems capable of understanding, deciding, and acting on the real world.\nTopics:\n① How can spatiotemporal data and multimodal foundation models support real-world cognition?\n② How can large models move from transportation knowledge question answering to planning and decision-making?\n③ How can agents enter the closed loop of transportation operation and control?\n④ How can the generalization, safety, and trustworthiness of autonomous systems be evaluated?","is_demo":0,"cover_url":"/api/forums/4212/cover?size=thumb","cover_alt":"How Can AI Enter the Real World? From Transportation Large Models and Agents to Autonomous Systems","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4212","source_checked_at":"2026-10-30","agenda_json":"[[\"Urban Transportation Spatiotemporal Intelligence: Cross-Scale, Cross-Domain, and Cross-Agent Perspectives\",\"汪炀\"],[\"Collaborative Adaptive Updating Methods for Satellite-Borne Models: Technical Support for Service Safety Monitoring of Transportation Infrastructure\",\"杨晨\"],[\"Transportation Spatiotemporal Large Models: From Specialized Prediction Models to General Spatiotemporal Reasoning\",\"万怀宇\"],[\"City-Level Intelligent Transportation Decision-Making: From Simulation and Reinforcement Learning to Real-World Deployment\",\"郑冠杰\"],[\"Human-Centered Intelligent Mobility: Interactive Experience and Diverse Population Needs\",\"李芳宇\"]]","code":"CNCC2026-4212","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"天府国际会议中心","room":"天府国际会议中心-207A天府厅A","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4212","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"人工智能如何走向真实世界？从交通大模型、智能体到自主系统","original_topic":"AI+行业","original_description":"当前人工智能正在由数字空间中的内容生成和知识问答，走向真实世界中的环境感知、任务规划、交互决策与自主运行。交通系统具有多源异构数据、复杂时空关联、动态开放环境、强实时性和安全约束，是检验大模型与智能体真实能力的重要场景。本论坛围绕交通大模型、智能体决策、世界模型、仿真评测、可信自主与端边云协同等问题，探讨人工智能如何从单点算法和辅助工具，演进为能够理解、决策并作用于真实世界的自主智能系统。\n议题：\n①时空数据与多模态基础模型如何支撑真实世界认知？\n②大模型如何由交通知识问答走向规划与决策？\n③智能体如何进入交通运行和控制闭环？\n④如何评测自主系统的泛化性、安全性与可信性？","people":[{"forum_id":4212,"role":"chair","name":"陈婧","affiliation":"杭州电子科技大学","sort_order":0},{"forum_id":4212,"role":"guest","name":"汪炀","affiliation":"中国科学技术大学","sort_order":0},{"forum_id":4212,"role":"co_chair","name":"曹原周汉","affiliation":"北京交通大学","sort_order":1},{"forum_id":4212,"role":"guest","name":"杨晨","affiliation":"西南交通大学","sort_order":1},{"forum_id":4212,"role":"guest","name":"万怀宇","affiliation":"北京交通大学","sort_order":2},{"forum_id":4212,"role":"guest","name":"郑冠杰","affiliation":"上海交通大学","sort_order":3},{"forum_id":4212,"role":"guest","name":"李芳宇","affiliation":"西南交通大学","sort_order":4}],"host":"","image_url":"/api/forums/4212/cover"},{"id":4213,"title":"Artificial Intelligence Methods for the Full Stack of Quantum Computing: Opportunities and Verifiability Challenges","topic":"Machine Learning Theory and Mechanisms","kind":"专题论坛","chair":"穆清、朱焱麟","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-104黄龙","description":"At the end of 2025, a review co-authored by 28 scholars in Nature Communications pointed out that AI is deeply intervening in the full stack of quantum computing from three levels: calibration, error correction, and circuit design. While AI moves from an \"auxiliary tool\" to a \"core method,\" it also brings new problems: how can the correctness and trustworthiness of quantum results produced or optimized by AI be verified? This forum sets up three topics around this main line.\nTopic 1: AI calibration pushes single-qubit fidelity to 99.9%: at the level of statistical confidence, can calibration methods scale with the number of qubits to support verification of thousand-qubit-scale systems?\nTopic 2: Resource allocation for quantum error correction: between near-term quantum error mitigation (QEM) and long-term quantum error correction (QEC), which path should AI prioritize empowering?\nTopic 3: AI circuit generation produces results with better gate counts but is difficult to understand: who guarantees correctness? AI-generated quantum circuits have better gate counts but are difficult to explain: how should functional correctness, equivalence verification, and attribution of responsibility be guaranteed?","is_demo":0,"cover_url":"/api/forums/4213/cover?size=thumb","cover_alt":"Artificial Intelligence Methods for the Full Stack of Quantum Computing: Opportunities and Verifiability Challenges","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4213","source_checked_at":"2026-10-30","agenda_json":"[[\"Toward Practical Quantum Artificial Intelligence\",\"张鹏\"],[\"Architecture Design of High-Performance qLDPC Decoding Accelerators\",\"卢丽强\"],[\"Hardware Architecture Outlook for High-Fidelity Low-Latency Quantum Error Correction\",\"王成\"],[\"Talk Topic: Design, Implementation, and Deployment of Real-Time Decoding Systems\",\"陈建鑫\"]]","code":"CNCC2026-4213","forum_type":"专题论坛","format":null,"theme_code":"机器学习理论与机理","theme_name":"机器学习理论与机理","venue":"天府国际会议中心","room":"天府国际会议中心-104黄龙","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4213","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"面向量子计算全栈的人工智能方法：机遇与可验证性挑战","original_topic":"机器学习理论与机理","original_description":"2025年底，Nature Communications 一篇28位学者联名的综述指出，AI正从标定、纠错与电路设计三个层面深度介入量子计算全栈。AI从\"辅助工具\"走向\"核心方法\"的同时，也带来了新的问题：由AI产生或优化的量子结果，其正确性与可信性如何验证？ 本论坛围绕这一主线设置三个议题。\n议题1：AI标定将单比特保真度推至99.9%：在统计置信度层面，标定方法能否随比特数扩展，支撑千比特级系统的验证？\n议题2：量子纠错的资源分配：近期量子错误缓解（QEM）与远期量子纠错（QEC）之间，AI应优先赋能哪条路径？\n议题3：AI线路生成，结果门数更优却难理解，正确性由谁来担保？AI生成的量子线路门数更优但难以解释：功能正确性、等价性验证与责任归属应如何保证？","people":[{"forum_id":4213,"role":"chair","name":"穆清","affiliation":"信息工程大学","sort_order":0},{"forum_id":4213,"role":"guest","name":"张鹏","affiliation":"天津大学","sort_order":0},{"forum_id":4213,"role":"co_chair","name":"朱焱麟","affiliation":"成都先进金属材料产业技术研究院股份有限公司","sort_order":1},{"forum_id":4213,"role":"guest","name":"韩永建","affiliation":"中国科学技术大学","sort_order":1},{"forum_id":4213,"role":"guest","name":"卢丽强","affiliation":"浙江大学","sort_order":2},{"forum_id":4213,"role":"guest","name":"王成","affiliation":"电子科技大学","sort_order":3},{"forum_id":4213,"role":"guest","name":"陈建鑫","affiliation":"清华大学","sort_order":4},{"forum_id":4213,"role":"guest","name":"郑盛根","affiliation":"粤港澳大湾区（广东）量子科学中心","sort_order":5},{"forum_id":4213,"role":"guest","name":"李晓瑜","affiliation":"电子科技大学/喀什地区电子信息产业技术研究院","sort_order":6},{"forum_id":4213,"role":"guest","name":"李绿周","affiliation":"中山大学","sort_order":7},{"forum_id":4213,"role":"guest","name":"张辉","affiliation":"本源量子计算科技（合肥）股份有限公司","sort_order":8},{"forum_id":4213,"role":"guest","name":"赵博","affiliation":"中国人民解放军信息工程大学","sort_order":9},{"forum_id":4213,"role":"guest","name":"孙回回","affiliation":"信息工程大学","sort_order":10}],"host":"","image_url":"/api/forums/4213/cover"},{"id":4222,"title":"2026 CCF Enterprise Digital Development Special Forum","topic":"AI for Industry","kind":"专题论坛","chair":"刘湘雯","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-110卧龙","description":"This forum focuses on \"AI reshaping a new paradigm for enterprise digitalization\", exploring how AI technology moves from concept to implementation and drives deep transformation in enterprise digitalization.\nThe forum will discuss in depth the challenges of AI engineering implementation, the reconstruction of organizations and businesses by AI-native applications, and the practice of frontier technologies such as embodied intelligence in vertical industries such as agriculture and animal husbandry. At the same time, it will directly address the \"AI divide\" and commercialization difficulties commonly faced by enterprises, and explore pragmatic paths from technical validation to a value closed loop.\nThis topic is recommended to respond to enterprises' urgent need for AI to move from \"usable\" to \"easy to use\", promote technology inclusiveness, and help enterprises cross the divide in the wave of intelligentization, achieving true cost reduction and efficiency improvement as well as model innovation.","is_demo":0,"cover_url":"/api/forums/4222/cover?size=thumb","cover_alt":"2026 CCF Enterprise Digital Development Special Forum","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4222","source_checked_at":"2026-10-30","agenda_json":"[[\"Host Opening Introduction\",\"谢婞敏\"],[\"Opening Remarks\",\"刘湘雯\"],[\"GAC Group's \\\"Data Weaving + Trusted Data Space + AI Empowerment\\\" Group-Level Data Intelligence Infrastructure Construction Project\",\"江泽鑫\"],[\"AI-Powered Chicken Farming: Wens Poultry's Comprehensive Digital Transformation Practice\",\"马铭龙\"],[\"Building a Solid Digital Foundation, Intelligence-Led Computing for the Future: Sichuan Bank's Digital Foundation and Digital-Intelligent Transformation Practice\",\"肖国彬\"],[\"Practice of an AI-Powered Smart Energy Management Platform in the Pharmaceutical Industry\",\"陈苏超\"],[\"The \\\"Chinese Ancient Books Data Resource Development and Application Platform\\\" Supporting the Development of the Ancient Book Digitization Industry\",\"罗志尊\"],[\"Panel: AI-Driven Industrial Upgrading\",\"刘湘明\"],[\"Concluding Remarks\",\"束庆山\"]]","code":"CNCC2026-4222","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"天府国际会议中心","room":"天府国际会议中心-110卧龙","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":9,"forum_id":"CNCC2026-4222","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"2026 CCF企业数字化发展专题论坛","original_topic":"AI+行业","original_description":"本论坛聚焦“AI重塑企业数字化新范式”，探讨AI技术如何从概念走向落地，驱动企业数字化转型的深层变革。\n论坛将深入讨论AI工程化落地的挑战、AI原生应用对组织与业务的重构、以及具身智能等前沿技术在农牧等垂直行业的实践。同时，直面企业普遍面临的“AI鸿沟”与商业化难题，探寻从技术验证到价值闭环的务实路径。\n推荐此选题，旨在回应企业对AI从“能用”到“好用”的迫切需求，推动技术普惠，助力企业在智能化浪潮中跨越鸿沟，实现真正的降本增效与模式创新。","people":[{"forum_id":4222,"role":"chair","name":"刘湘雯","affiliation":"阿里云智能集团","sort_order":0},{"forum_id":4222,"role":"guest","name":"谢婞敏","affiliation":"阿里云智能集团","sort_order":0},{"forum_id":4222,"role":"guest","name":"江泽鑫","affiliation":"广州汽车集团股份有限公司","sort_order":1},{"forum_id":4222,"role":"guest","name":"马铭龙","affiliation":"温氏食品集团股份有限公司","sort_order":2},{"forum_id":4222,"role":"guest","name":"肖国彬","affiliation":"四川银行","sort_order":3},{"forum_id":4222,"role":"guest","name":"陈苏超","affiliation":"华润三九医药股份有限公司","sort_order":4},{"forum_id":4222,"role":"guest","name":"罗志尊","affiliation":"北京古集信息科技有限公司","sort_order":5},{"forum_id":4222,"role":"guest","name":"刘湘明","affiliation":"钛媒体","sort_order":6},{"forum_id":4222,"role":"guest","name":"左红","affiliation":"四川长虹电器股份有限公司","sort_order":7},{"forum_id":4222,"role":"guest","name":"蒋林玻","affiliation":"重庆蚂蚁消费金融有限公司","sort_order":8},{"forum_id":4222,"role":"guest","name":"李琨","affiliation":"成都先进金属材料产业技术研究院股份有限公司","sort_order":9},{"forum_id":4222,"role":"guest","name":"殷国良","affiliation":"新疆天业(集团)有限公司","sort_order":10},{"forum_id":4222,"role":"guest","name":"常旭岭","affiliation":"宁波书写芯忆科技有限公司","sort_order":11},{"forum_id":4222,"role":"guest","name":"束庆山","affiliation":"中国计算机学会业务总部","sort_order":12}],"host":"","image_url":"/api/forums/4222/cover"},{"id":4233,"title":"Youth Growth Forum: Academic Community Empowering Young Sci-Tech Workers' Growth Paths and Value Reconstruction","topic":"General","kind":"专题论坛","chair":"唐诗、张忠宝","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-509","description":"Young sci-tech workers are facing unprecedented multiple challenges—a monolithic research evaluation system, narrowing career development channels, intensifying academic involution, and misaligned social expectations—and their personal growth paths are becoming increasingly blurred. This forum invites outstanding young representatives from the national YOCSEF system who have graduated and grown up, together with the Youth Working Committee, to focus on the growth of young sci-tech workers.\nDebate Topic 1 · Facing the Predicament (Looking Inward):\nUnder a diversified evaluation system, how can young faculty achieve self-development and break through in value?\nDebate Topic 2 · Grasping the Changing Landscape (Looking Outward):\nAmid the AI wave, how can young faculty reshape their competitiveness amid opportunities and challenges?\nDebate Topic 3 · Community Empowerment (Looking Upward):\nHow can the academic community truly become a breakthrough force for youth growth?","is_demo":0,"cover_url":"/api/forums/4233/cover?size=thumb","cover_alt":"Youth Growth Forum: Academic Community Empowering Young Sci-Tech Workers' Growth Paths and Value Reconstruction","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4233","source_checked_at":"2026-10-30","agenda_json":"[[\"Strategy, Innovation, and Collaboration: The Growth of Young Scholars in the Intelligent Era\",\"李建欣\"],[\"How to Build Stepping Stones for Research Work to \\\"Break Out\\\"?\",\"包云岗\"]]","code":"CNCC2026-4233","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城5号馆","room":"西博城5号馆-509","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":2,"forum_id":"CNCC2026-4233","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"青年成长论坛-学术共同体赋能青年科技工作者成长路径与价值重构","original_topic":"综合","original_description":"青年科技工作者正面临前所未有的多重挑战——科研评价体系单一化、职业发展通道收窄、学术内卷加剧与社会期待错位，个人成长路径日趋模糊。本论坛邀请全国YOCSEF体系中已毕业并成长起来的优秀青年代表，联合青工委，聚焦青年科技工作者成长。\n思辨议题一 · 直面困局（向内看）：\n多元评价体系下，青年教师如何实现自我发展与价值突围？\n思辨议题二 · 把握变局（向外看）：\nAI 浪潮之下，青年教师如何在机遇与挑战中重塑竞争力？\n思辨议题三 · 共同体赋能（向上看）：\n学术共同体如何真正成为青年成长的破局力量？","people":[{"forum_id":4233,"role":"guest","name":"李建欣","affiliation":"北京航空航天大学","sort_order":0},{"forum_id":4233,"role":"chair","name":"唐诗","affiliation":"真术相成（成都）科技有限公司","sort_order":0},{"forum_id":4233,"role":"co_chair","name":"张忠宝","affiliation":"北京邮电大学","sort_order":1},{"forum_id":4233,"role":"guest","name":"黄栋","affiliation":"华南农业大学","sort_order":1},{"forum_id":4233,"role":"guest","name":"包云岗","affiliation":"中国科学院计算技术研究所","sort_order":2},{"forum_id":4233,"role":"guest","name":"吴天星","affiliation":"东南大学","sort_order":3},{"forum_id":4233,"role":"guest","name":"况琨","affiliation":"浙江大学","sort_order":4},{"forum_id":4233,"role":"guest","name":"王宝生","affiliation":"国防科技大学","sort_order":5},{"forum_id":4233,"role":"guest","name":"张超","affiliation":"清华大学","sort_order":6}],"host":"","image_url":"/api/forums/4233/cover"},{"id":4236,"title":"The \"Intelligence\" and \"Emotion\" of Video Generation","topic":"Vision, Multimodality and World Models","kind":"专题论坛","chair":"杨巨峰","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城5号馆-502","description":"In 2024, \"Research on Digital Humans and Robots with Both Emotion and Intelligence\" was ranked first among the Top Ten Frontier Scientific Questions of the China Association for Science and Technology. Inspired by this, this forum takes the \"intelligence\" and \"emotion\" of video generation as its theme, exploring how intelligent algorithms understand the world and how the warmth of emotion can be infused into rational computation. \"Intelligence\" points to technological breakthroughs in multimodal large models, physical simulation, and spatiotemporal modeling, which are the cornerstone of realism; \"emotion\" concerns the shaping of narrative rhythm, aesthetic imagery, and empathy, which is the source of appeal. We believe that true video intelligence goes beyond generating images—it lies in touching people's hearts. The forum will bring together technical experts and creative practitioners to jointly ask: how can we find a balance between efficiency and warmth, between realism and expressiveness, so that video generation technology is not only smarter but also understands us better. We sincerely invite you to join this in-depth dialogue about the future of imagery.","is_demo":0,"cover_url":"/api/forums/4236/cover?size=thumb","cover_alt":"The \"Intelligence\" and \"Emotion\" of Video Generation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4236","source_checked_at":"2026-10-30","agenda_json":"[[\"Emotionally and Cognitively Capable Multi-Agent Content Generation and Social Simulation\",\"刘武\"],[\"Multi-Agent Collaborative Long-Sequence Video Generation: From Rule Controllability to Narrative Empathy\",\"户保田\"],[\"How Visual Generation and Understanding Move Toward Collaborative Unification\",\"袁粒\"],[\"Physical Awareness and Audio-Visual Emotional Expression in Video Generation\",\"亓帆\"],[\"From Video Generation to Cognitive Influence: Multimodal Content Computing for Cognitive Security\",\"折栋宇\"]]","code":"CNCC2026-4236","forum_type":"专题论坛","format":null,"theme_code":"视觉、多模态与世界模型","theme_name":"视觉、多模态与世界模型","venue":"西博城5号馆","room":"西博城5号馆-502","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4236","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"视频生成的“智”与“情”","original_topic":"视觉、多模态与世界模型","original_description":"2024年，“情智兼备的数字人与机器人的研究”入选中国科协十大前沿科学问题第1位。受此启发，本次论坛以视频生成的“智”与“情”为主题，探讨智能算法如何理解世界，以及如何在理性计算中注入情感的温度。“智”指向多模态大模型、物理仿真与时空建模的技术突破，是真实感的基石；“情”则关乎叙事节奏、审美意象与共情能力的塑造，是感染力的源泉。我们相信，真正的视频智能不止于生成画面，更在于触动人心。论坛将汇聚技术专家与创意实践者，共同追问：如何在效率与温度、写实与写意之间找到平衡，让视频生成技术不仅更聪明，也更懂我们。诚邀你一同走进这场关于未来影像的深度对话。","people":[{"forum_id":4236,"role":"chair","name":"杨巨峰","affiliation":"南开大学","sort_order":0},{"forum_id":4236,"role":"guest","name":"刘武","affiliation":"中国科学技术大学","sort_order":0},{"forum_id":4236,"role":"guest","name":"户保田","affiliation":"哈尔滨工业大学（深圳）","sort_order":1},{"forum_id":4236,"role":"guest","name":"袁粒","affiliation":"北京大学","sort_order":2},{"forum_id":4236,"role":"guest","name":"亓帆","affiliation":"内蒙古大学","sort_order":3},{"forum_id":4236,"role":"guest","name":"折栋宇","affiliation":"中关村实验室","sort_order":4}],"host":"","image_url":"/api/forums/4236/cover"},{"id":4248,"title":"Software Engineering Discipline Construction and Top Innovative Talent Cultivation in the AI Era","topic":"General","kind":"专题论坛","chair":"张小松","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"天府国际会议中心-103剑门关","description":"With the development of large models and artificial intelligence technologies, the talent cultivation paradigm and curriculum system construction of software engineering have been greatly affected. Facing the \"penetrating\" impact of artificial intelligence, traditional talent cultivation models urgently need systematic transformation, should focus on cultivating students' computational thinking and ability to solve complex engineering problems, and at the same time actively explore innovative paths for interdisciplinary integration and personalized cultivation of top talents. This forum application, as an important part of the CNCC2026 cooperative enterprise forum, will join hands with top universities, research institutes, and leading industry enterprises to deeply discuss the paths for software engineering discipline construction and top innovative talent cultivation in the AI era.","is_demo":0,"cover_url":"/api/forums/4248/cover?size=thumb","cover_alt":"Software Engineering Discipline Construction and Top Innovative Talent Cultivation in the AI Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4248","source_checked_at":"2026-10-30","agenda_json":"[[\"Challenges and Insights for Software Engineering Education in the AI Era\",\"王建民\"],[\"Cultivating Outstanding Engineers in Key Software Domains\",\"胡春明\"],[\"Exploring Reform of Software School Management and Operating Mechanisms in the AI Era\",\"张旭鸿\"],[\"Practice and Challenges in Cultivating Open Source and Xinchuang Software Talent\",\"玄跻峰\"],[\"Strengthening Dual Foundations, Building Platforms, and Innovating Models to Cultivate Software Security Innovators\",\"廖勇\"]]","code":"CNCC2026-4248","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"天府国际会议中心","room":"天府国际会议中心-103剑门关","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4248","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI时代软件工程学科建设与拔尖创新人才培养","original_topic":"综合","original_description":"随着大模型和人工智能技术的发展，软件工程人才培养范式与课程体系建设受到极大的影响。面对人工智能的“穿透性”影响，传统人才培养模式亟待系统性转型，应着力培养学生计算思维与复杂工程问题解决能力，同时积极探索跨学科融合与拔尖人才个性化培养的创新路径。本次论坛申请作为CNCC2026合作企业论坛的重要组成部分，将联合顶尖高校、科研院所及行业领军企业，深度探讨AI时代软件工程学科建设和拔尖创新人才培养路径。","people":[{"forum_id":4248,"role":"chair","name":"张小松","affiliation":"电子科技大学","sort_order":0},{"forum_id":4248,"role":"guest","name":"王建民","affiliation":"清华大学","sort_order":0},{"forum_id":4248,"role":"guest","name":"胡春明","affiliation":"北京航空航天大学","sort_order":1},{"forum_id":4248,"role":"guest","name":"张旭鸿","affiliation":"浙江大学","sort_order":2},{"forum_id":4248,"role":"guest","name":"玄跻峰","affiliation":"武汉大学","sort_order":3},{"forum_id":4248,"role":"guest","name":"廖勇","affiliation":"电子科技大学","sort_order":4}],"host":"","image_url":"/api/forums/4248/cover"},{"id":4249,"title":"CCF Outstanding Doctoral Dissertation Forum: Future Development of AI and Talent Cultivation","topic":"General","kind":"专题论坛","chair":"俞扬","starts_at":"2026-10-23T13:30:00+08:00","ends_at":"2026-10-23T17:30:00+08:00","place":"西博城4号馆-415","description":"CCF has established the Outstanding Doctoral Dissertation Award since 2006, annually selecting no more than 10 CCF Outstanding Doctoral Dissertation recipients (abbreviated as CCF Outstanding Doctoral Dissertations) from doctoral graduates in computer science at domestic universities. Over more than a decade, more than one hundred doctoral graduates have received this honor. The recipients of CCF Outstanding Doctoral Dissertations have gradually grown into internationally renowned university professors, Changjiang Scholars, National Science Fund for Distinguished Young Scholars and Excellent Young Scholars, and corporate general managers/directors, among others. This forum invites some outstanding representatives from past CCF Outstanding Doctoral Dissertation recipients, who will share their outlook on the future development of artificial intelligence and jointly discuss the challenges and breakthrough opportunities in talent cultivation in the era of artificial intelligence, to support the cultivation and development of high-level computer science doctoral students in China.","is_demo":0,"cover_url":"/api/forums/4249/cover?size=thumb","cover_alt":"CCF Outstanding Doctoral Dissertation Forum: Future Development of AI and Talent Cultivation","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4249","source_checked_at":"2026-10-30","agenda_json":"[[\"Online Learning and Large Model Inference Optimization\",\"赵鹏\"]]","code":"CNCC2026-4249","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城4号馆","room":"西博城4号馆-415","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":1,"forum_id":"CNCC2026-4249","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"CCF优博论坛：人工智能的未来发展与人才培养","original_topic":"综合","original_description":"CCF从2006年起设立优秀博士学位论文奖，每年从国内高校计算机学科博士毕业生中，评选不超过10名CCF优秀博士学位论文获得者（简称CCF优博），十余年已有一百多位博士获此殊荣。CCF优博的获奖者逐渐成长为国际知名大学教授、长江学者、国家杰青优青、企业总经理/总监等。本论坛从历届CCF优博获得者中邀请部分杰出代表，将分享对人工智能未来发展的展望，并共同讨论在人工智能时代人才培养的挑战与突破机遇，助力我国高水平计算机博士研究生的培养和发展。","people":[{"forum_id":4249,"role":"chair","name":"俞扬","affiliation":"南京大学","sort_order":0},{"forum_id":4249,"role":"guest","name":"章国锋","affiliation":"浙江大学","sort_order":0},{"forum_id":4249,"role":"guest","name":"刘亮","affiliation":"北京邮电大学","sort_order":1},{"forum_id":4249,"role":"guest","name":"李泽超","affiliation":"南京理工大学","sort_order":2},{"forum_id":4249,"role":"guest","name":"李崇轩","affiliation":"中国人民大学","sort_order":3},{"forum_id":4249,"role":"guest","name":"赵鹏","affiliation":"南京大学","sort_order":4}],"host":"","image_url":"/api/forums/4249/cover"},{"id":4299,"title":"Plenary Forum (Oct 24) — What and How Should We Learn and Work in the AI Era?","topic":"Unclassified","kind":"大会论坛","chair":"文继荣（主持）","starts_at":"2026-10-24T10:20:00+08:00","ends_at":"2026-10-24T11:20:00+08:00","place":"西博城1号馆","description":"As artificial intelligence becomes increasingly integrated into learning and work, what should we learn, and what should we do? How will the value of knowledge, academic credentials, and experience change? How should the opportunities and transformation costs brought by technological progress be shared and borne? This roundtable forum invites experts in artificial intelligence, education, and employment to start from the capabilities and boundaries of technology and discuss the meaning of learning, the transformation of talent cultivation, new paths for career development, and human autonomy and creativity. This one-hour cross-disciplinary dialogue will provide students, educators, and working professionals with reflections and inspiration for understanding changes and making choices.","is_demo":0,"cover_url":"/api/forums/4299/cover?size=thumb","cover_alt":"Plenary Forum (Oct 24) — What and How Should We Learn and Work in the AI Era?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4299","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4299","forum_type":"大会论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城1号馆","room":"西博城1号馆","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[\"官网标注主持，未标注主席；主持与主席分开存储。\"]","agenda_count":0,"forum_id":"CNCC2026-4299","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"24日大会论坛——人工智能时代，我们学什么、干什么？","original_topic":"未分类","original_description":"当人工智能日益融入学习与工作，我们该学什么，又该干什么？知识、学历与经验的价值将如何变化？技术进步带来的机会与转型成本，又该如何分享与承担？本圆桌论坛邀请人工智能、教育与就业领域的专家，从技术的能力与边界出发，探讨学习的意义、人才培养的转型、职业发展的新路径，以及人的自主性与创造力等。一小时的跨界对话，将为学生、教育工作者与职场人士提供理解变化、作出选择的思考与启发。","people":[{"forum_id":4299,"role":"host","name":"文继荣","affiliation":"中国人民大学","sort_order":0},{"forum_id":4299,"role":"guest","name":"包云岗","affiliation":"中国科学院计算技术研究所","sort_order":0},{"forum_id":4299,"role":"guest","name":"陈文光","affiliation":"清华大学计算机科学与技术系","sort_order":1},{"forum_id":4299,"role":"guest","name":"章文嵩","affiliation":"安托盟丘","sort_order":2},{"forum_id":4299,"role":"guest","name":"赵忠","affiliation":"中国人民大学","sort_order":3}],"host":"文继荣","image_url":"/api/forums/4299/cover"},{"id":4009,"title":"New Methods and Paradigms for AI-Empowered Chip Design","topic":"AI for Industry","kind":"专题论坛","chair":"贾天宇","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城2号馆-205","description":"As chip design complexity and tape-out costs rise exponentially, traditional design methods centered on empirical rules and manual iteration are facing unprecedented challenges. The rise of large models, especially agent technologies, has opened up an entirely new paradigm for chip design automation, and AI is gradually permeating every stage from architecture generation to physical implementation. This forum has invited multiple authoritative experts from universities and industry, focusing on frontier explorations of AI technologies such as AI Agents in empowering new paradigms of chip design. Through the intellectual exchange at the forum, we look forward to systematically reviewing the current progress, bottlenecks, and future roadmap of intelligent chip design, and to providing useful references for promoting interdisciplinary integration and innovative breakthroughs in the field.","is_demo":0,"cover_url":"/api/forums/4009/cover?size=thumb","cover_alt":"New Methods and Paradigms for AI-Empowered Chip Design","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4009","source_checked_at":"2026-10-30","agenda_json":"[[\"Controllable Intervention Points for LLMs in Digital Chip Design: Methods, Boundaries, and Expectations\",\"邸志雄\"],[\"2.5D/3D Agent Acceleration Architectures and AI-Assisted Design Methods\",\"贾天宇\"],[\"AI-Driven Chip Verification: Practice of the UCAgent Agent in the XiangShan Processor\",\"姚治成\"],[\"Breaking the Speed Limits of DDR and SerDes: Building High-Bandwidth Memory and High-Speed Interconnect Engines for AI Large Models\",\"王刚\"],[\"Co-evolution of AI and EDA: S2C Digital EDA Empowers Large-Scale Chip Innovation\",\"梁建忠\"]]","code":"CNCC2026-4009","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"西博城2号馆","room":"西博城2号馆-205","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4009","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI赋能芯片设计新方法与新模式","original_topic":"AI+行业","original_description":"随着芯片设计的复杂度与流片成本呈指数级上升，传统以经验规则和人工迭代为主的设计方法正面临前所未有的挑战。大模型，特别是智能体技术的兴起，为芯片设计自动化开辟了全新范式，AI正逐步渗透从架构生成到物理实现的各个环节。本论坛邀请到来自高校及产业界的多位权威专家，聚焦AI Agent等人工智能技术在赋能芯片设计新范式的前沿探索。期待通过论坛的思维碰撞，系统梳理智能化芯片设计的当前进展、瓶颈与未来路线图，为推动领域交叉与创新突破提供有益参考。","people":[{"forum_id":4009,"role":"chair","name":"贾天宇","affiliation":"北京大学","sort_order":0},{"forum_id":4009,"role":"guest","name":"邸志雄","affiliation":"西南交通大学","sort_order":0},{"forum_id":4009,"role":"guest","name":"贾天宇","affiliation":"北京大学","sort_order":1},{"forum_id":4009,"role":"guest","name":"姚治成","affiliation":"中国科学院计算技术研究所","sort_order":2},{"forum_id":4009,"role":"guest","name":"王刚","affiliation":"芯动微电子科技（北京）有限公司","sort_order":3},{"forum_id":4009,"role":"guest","name":"梁建忠","affiliation":"上海思尔芯技术股份有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4009/cover"},{"id":4011,"title":"Spatiotemporal Federated Learning: From Data Islands to Intelligent Collaboration","topic":"Machine Learning Theory and Mechanisms","kind":"专题论坛","chair":"桂佳平","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-406","description":"This forum focuses on spatiotemporal federated learning in IoT and edge intelligence scenarios, discusses core challenges such as data heterogeneity, privacy protection, and model security, and explores new paradigms of cross-domain trustworthy sharing and spatiotemporal continual learning. The forum conducts in-depth exchanges on frontier topics such as federated inference optimization, heterogeneous data adaptation, continual learning collaboration, and privacy and attack resistance, providing new insights from both academic frontiers and industrial implementation perspectives for building an efficient, secure, and sustainable next-generation collaborative intelligence system.","is_demo":0,"cover_url":"/api/forums/4011/cover?size=thumb","cover_alt":"Spatiotemporal Federated Learning: From Data Islands to Intelligent Collaboration","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4011","source_checked_at":"2026-10-30","agenda_json":"[[\"Research on Key Technologies for Federated Learning and Inference Optimization for IoT Applications\",\"杜军朝\"],[\"Challenges and Evolution of Federated Learning: Toward Efficient Collaboration and Secure Trustworthiness\",\"王志波\"],[\"Exploration of Heterogeneous Federated Learning for Cross-Domain Trusted Data Sharing\",\"盖珂珂\"],[\"Privacy Protection in Federated Learning for Multi-Source Heterogeneous Environments\",\"苏洲\"],[\"A New Paradigm for Multi-Agent Swarm Intelligence Collaboration: Theory, Technology, and Applications of Federated Continual Learning\",\"杨新\"]]","code":"CNCC2026-4011","forum_type":"专题论坛","format":null,"theme_code":"机器学习理论与机理","theme_name":"机器学习理论与机理","venue":"西博城4号馆","room":"西博城4号馆-406","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4011","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"时空联邦学习：从数据孤岛到智能协同","original_topic":"机器学习理论与机理","original_description":"本论坛聚焦物联网与边缘智能场景下的时空联邦学习，探讨数据异构、隐私保护与模型安全等核心挑战，探索跨域可信共享与时空持续学习新范式。论坛围绕联邦推理优化、异构数据适配、持续学习协同、隐私抗攻击等前沿议题展开深入交流，从学术前沿与产业落地双重视角，为构建高效、安全、可持续的下一代协同智能体系提供新洞见。","people":[{"forum_id":4011,"role":"chair","name":"桂佳平","affiliation":"上海交通大学","sort_order":0},{"forum_id":4011,"role":"guest","name":"杜军朝","affiliation":"西安电子科技大学计算机科学与技术学院","sort_order":0},{"forum_id":4011,"role":"guest","name":"王志波","affiliation":"浙江大学","sort_order":1},{"forum_id":4011,"role":"guest","name":"盖珂珂","affiliation":"北京理工大学","sort_order":2},{"forum_id":4011,"role":"guest","name":"苏洲","affiliation":"西安交通大学","sort_order":3},{"forum_id":4011,"role":"guest","name":"杨新","affiliation":"西南财经大学","sort_order":4}],"host":"","image_url":"/api/forums/4011/cover"},{"id":4017,"title":"Integrated Memory Systems for Embodied Intelligence and AI Agents","topic":"Unclassified","kind":"专题论坛","chair":"宋继强","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-509","description":"Memory, as the core of artificial intelligence, has become increasingly prominent in status and importance with the rapid and in-depth development of AI agents and embodied intelligence. This forum will focus on integrated memory systems for embodied intelligence and AI agents, exploring from different levels and perspectives the frontier research, technical challenges, practical applications, and deployment cases of memory systems, including in-depth analysis and exchange on system architecture, performance bottlenecks, security and privacy, and future development, so as to spark new inspiration and continuously promote frontier innovation and practical deployment of memory systems.","is_demo":0,"cover_url":"/api/forums/4017/cover?size=thumb","cover_alt":"Integrated Memory Systems for Embodied Intelligence and AI Agents","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4017","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Remarks\",\"宋继强\"],[\"Making Robots Move and Think Like Humans\",\"仉尚航\"],[\"Latest Practices in Robotics and Industry Intelligence\",\"毛世杰\"],[\"Industrial-Grade General Agent Memory: Marvis Memory System Architecture Design and Practice\",\"蔡建涛\"],[\"Long-Term Memory for Embodied Agents Integrating Symbolic Methods and Large Models\",\"张益民\"],[\"The Impossible Triangle of Embodied Memory: Heterogeneous System Design Under Latency, Energy Efficiency, and Privacy\",\"吴向斌\"]]","code":"CNCC2026-4017","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-509","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4017","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"具身智能与智能体的融合记忆系统","original_topic":"未分类","original_description":"记忆作为人工智能的核心，随着智能体和具身智能的高速和深入发展，其地位和重要性愈发凸显。本论坛将围绕具身智能以及智能体的融合记忆系统展开，从不同的层次和角度，探讨记忆系统的前沿研究、技术挑战、实际应用以及落地案例，包含系统架构、性能瓶颈、安全与隐私以及未来发展等方方面面的深入剖析和碰撞，出发新的灵感，持续推动记忆系统的前沿创新和实际落地。","people":[{"forum_id":4017,"role":"guest","name":"仉尚航","affiliation":"北京大学","sort_order":0},{"forum_id":4017,"role":"chair","name":"宋继强","affiliation":"英特尔中国研究院","sort_order":0},{"forum_id":4017,"role":"guest","name":"毛世杰","affiliation":"联想研究院","sort_order":1},{"forum_id":4017,"role":"guest","name":"蔡建涛","affiliation":"腾讯","sort_order":2},{"forum_id":4017,"role":"guest","name":"张益民","affiliation":"眸深智能","sort_order":3},{"forum_id":4017,"role":"guest","name":"吴向斌","affiliation":"英特尔中国研究院","sort_order":4}],"host":"","image_url":"/api/forums/4017/cover"},{"id":4019,"title":"Are Large Models the Watt Steam Engine? Reflections on Efficiency, Reliability, and Scaled Application","topic":"AI Foundation Models","kind":"专题论坛","chair":"欧智坚、宗成庆","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-415","description":"Artificial intelligence is accelerating into the core processes of industry and becoming an important driving force for new industrialization and the transformation of the digital economy. Large models represented by ChatGPT, DeepSeek, Tongyi Qianwen, and others enable machines to participate in knowledge production, software development, engineering design, and operations management through natural language, multimodal interaction, and tool invocation. Analogous to how the Watt steam engine achieved a leap in efficiency and stability through the separate condenser and expanded from mine drainage to textiles, metallurgy, transportation, and other fields, do today's large models already possess the efficiency, reliability, and scaled application capabilities to support a new industrial revolution? If not, what is this \"Watt-style breakthrough\": general foundation models, long-horizon reliable agents, specialized domain models, unified multimodal models, collaboration between large and small models, or new intelligent infrastructure for Token production and scheduling? This forum will use a deliberative format to distill judgment frameworks and consensus on industrial applications.","is_demo":0,"cover_url":"/api/forums/4019/cover?size=thumb","cover_alt":"Are Large Models the Watt Steam Engine? Reflections on Efficiency, Reliability, and Scaled Application","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4019","source_checked_at":"2026-10-30","agenda_json":"[[\"Trustworthy AI: Transforming Intelligent Systems into a Safe Industrial Engine\",\"冯俊兰\"],[\"Industry Transformation Brought by Large Models: An AI for Math Perspective\",\"王立威\"],[\"AI Large Models: From the Digital World to the Physical World\",\"何晓冬\"],[\"Research on Reasoning and Generalization of Language Models\",\"张岳\"],[\"Language Intelligence-Centered Multimodal Large Models and Human-Computer Interaction\",\"户保田\"]]","code":"CNCC2026-4019","forum_type":"专题论坛","format":null,"theme_code":"AI基础模型","theme_name":"AI基础模型","venue":"西博城4号馆","room":"西博城4号馆-415","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4019","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"大模型是瓦特蒸汽机吗？效率、可靠性与规模化应用的思辨","original_topic":"AI基础模型","original_description":"人工智能正加速进入产业核心流程，成为新型工业化、数字经济变革的重要驱动力。以ChatGPT、DeepSeek、通义千问等为代表的大模型，让机器以自然语言、多模态交互和工具调用方式参与知识生产、软件开发、工程设计与运营管理。类比瓦特蒸汽机通过分离冷凝器实现效率和稳定性跃迁，并从矿井排水走向纺织、冶金、交通运输等领域，今天的大模型是否已具备支撑新工业革命的效率、可靠性和规模化应用能力？如果还没有，这个“瓦特式突破”是什么：通用基础模型、长程可靠智能体、专业化垂域模型、多模态统一模型、大小模型协同，还是面向Token生产与调度的新型智能基础设施？本论坛将以思辨形式凝练判断框架与产业应用共识。","people":[{"forum_id":4019,"role":"chair","name":"欧智坚","affiliation":"清华大学","sort_order":0},{"forum_id":4019,"role":"guest","name":"冯俊兰","affiliation":"中国移动九天研究院","sort_order":0},{"forum_id":4019,"role":"guest","name":"王立威","affiliation":"北京大学","sort_order":1},{"forum_id":4019,"role":"co_chair","name":"宗成庆","affiliation":"中国科学院自动化研究所","sort_order":1},{"forum_id":4019,"role":"guest","name":"何晓冬","affiliation":"京东集团","sort_order":2},{"forum_id":4019,"role":"guest","name":"张岳","affiliation":"西湖大学","sort_order":3},{"forum_id":4019,"role":"guest","name":"户保田","affiliation":"哈尔滨工业大学（深圳）","sort_order":4}],"host":"","image_url":"/api/forums/4019/cover"},{"id":4039,"title":"Forum on Graph Computing Applications and Software-Hardware Support Technologies in the Intelligent Computing Era","topic":"Unclassified","kind":"专题论坛","chair":"张宇","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-401","description":"As an important tool for analyzing correlations among entities in the intelligent computing era, graph computing has been widely applied in numerous fields such as social governance, healthcare, power grid analysis, computational materials, computational breeding, and EDA. However, in recent years, with the vigorous development of artificial intelligence technologies, new graph computing paradigms such as graph foundation models have emerged, and demands have become increasingly complex and diverse. This has brought enormous challenges to existing computer architectures. How to design novel energy-efficient graph computing architectures and corresponding system software to meet complex and diverse graph computing demands is an urgent problem to be solved. To this end, this forum will discuss how to design energy-efficient architectures and system software for new graph computing applications, efficiently support complex graph computing demands such as graph foundation models, and help the development of graph computing technology and the maturation of its ecosystem.","is_demo":0,"cover_url":"/api/forums/4039/cover?size=thumb","cover_alt":"Forum on Graph Computing Applications and Software-Hardware Support Technologies in the Intelligent Computing Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4039","source_checked_at":"2026-10-30","agenda_json":"[[\"TianheGraph: Tianhe Graph Computing and Applications\",\"甘新标\"],[\"Design and Optimization of Large-Scale Graph Computing Systems for Serverless Cloud Platforms\",\"孙世轩\"],[\"A High-Level Synthesis Programming Model for Graph Computing Accelerators\",\"卓有为\"],[\"Domestic Computing Power Optimization: From RL Post-Training to Multimodal Generative Inference\",\"曾立\"],[\"Research on High-Performance Temporal Graph Analysis Techniques\",\"赵进\"]]","code":"CNCC2026-4039","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-401","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4039","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智能计算时代图计算应用及软硬件支撑技术论坛","original_topic":"未分类","original_description":"图计算作为智能计算时代分析事物之间关联关系的重要工具，已广泛地应用于社会治理、医疗健康、电网分析、计算材料、计算育种、EDA等众多领域。然而，近年随着人工智能技术蓬勃发展，图基础模型等新型图计算范式涌现，需求日益复杂多样。这给现有计算机体系结构带来了巨大挑战。如何设计新型高能效图计算体系结构和相应系统软件满足复杂多样的图计算需求是一个亟待解决的难题。为此，本论坛将讨论如何设计面向新型图计算应用的高能效体系结构及系统软件，高效支撑图基础模型等复杂图计算需求，助力图计算技术发展和生态成熟。","people":[{"forum_id":4039,"role":"chair","name":"张宇","affiliation":"华中科技大学","sort_order":0},{"forum_id":4039,"role":"guest","name":"甘新标","affiliation":"国防科技大学","sort_order":0},{"forum_id":4039,"role":"guest","name":"孙世轩","affiliation":"上海交通大学计算机学院","sort_order":1},{"forum_id":4039,"role":"guest","name":"卓有为","affiliation":"北京大学","sort_order":2},{"forum_id":4039,"role":"guest","name":"曾立","affiliation":"深圳市大数据研究院","sort_order":3},{"forum_id":4039,"role":"guest","name":"赵进","affiliation":"华中科技大学","sort_order":4}],"host":"","image_url":"/api/forums/4039/cover"},{"id":4052,"title":"Reinforcement Learning and Self-Evolution of Large-Model Agents","topic":"Reinforcement Learning and Reasoning","kind":"专题论坛","chair":"周明","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-409","description":"Currently, large-model agents are moving from traditional supervised training toward autonomous evolution, achieving significant progress in scenarios such as autonomous scientific discovery, software development, deep search, and data analysis. However, applying agents in real-world scenarios requires them to have continuous learning and self-evolution capabilities to achieve domain adaptation and meet user needs. This forum invites several outstanding scholars active on the front lines to share the cutting-edge technologies and future development trends of large-model agent self-evolution, discuss the technical challenges and technical approaches faced by agents in self-evolution in real-world scenarios, and discuss the development trends of agent self-evolution from the perspectives of agentic RL and agent harness as well as the impact on industry and AI products, promoting further application of agents in broader real-world scenarios.","is_demo":0,"cover_url":"/api/forums/4052/cover?size=thumb","cover_alt":"Reinforcement Learning and Self-Evolution of Large-Model Agents","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4052","source_checked_at":"2026-10-30","agenda_json":"[[\"Development and Challenges of RSI Recursive Self-Improvement\",\"周明\"],[\"Co-Evolving AI and Inexplicable yet Verifiable Innovation\",\"沈阳\"],[\"From Reinforcement Learning to Self-Evolution: Toward Long-Horizon Agents\",\"林衍凯\"],[\"The Evolution Recipe for Agentic AI: Feedback Loops of Models, Environments, and Data\",\"林涛\"],[\"Intelligent Evolution from a First-Principles Perspective: From Model Foundations to Runtime Frameworks\",\"钱泓锦\"]]","code":"CNCC2026-4052","forum_type":"专题论坛","format":null,"theme_code":"强化学习与推理","theme_name":"强化学习与推理","venue":"西博城4号馆","room":"西博城4号馆-409","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4052","legacy_id":6,"lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"大模型智能体强化学习与自进化","original_topic":"强化学习与推理","original_description":"当前，大模型智能体正从传统的监督训练走向自主进化，在自主科学发现、软件开发、深度搜索、数据分析等场景取得显著的进展。然而将智能体应用在真实场景中要求其具有持续学习与自进化能力，实现领域适配，满足用户需求。本论坛邀请多位活跃在一线的优秀学者分享大模型智能体自进化的前沿技术与未来发展趋势，探讨智能体在真实场景自进化所面临的技术挑战与技术思路，从agentic RL与agent harness角度讨论智能体自进化的发展趋势及对业界和AI产品带来的影响，推动智能体在更广泛真实场景的进一步应用。","people":[{"forum_id":4052,"role":"guest","name":"杨健","affiliation":"北京航空航天大学","sort_order":0},{"forum_id":4052,"role":"chair","name":"周明","affiliation":"北京澜舟科技有限公司","sort_order":0},{"forum_id":4052,"role":"guest","name":"陈冠华","affiliation":"南方科技大学","sort_order":1},{"forum_id":4052,"role":"guest","name":"沈阳","affiliation":"清华大学","sort_order":2},{"forum_id":4052,"role":"guest","name":"林衍凯","affiliation":"中国人民大学","sort_order":3},{"forum_id":4052,"role":"guest","name":"林涛","affiliation":"西湖大学","sort_order":4},{"forum_id":4052,"role":"guest","name":"钱泓锦","affiliation":"北京智源人工智能研究院","sort_order":5},{"forum_id":4052,"role":"guest","name":"王昊奋","affiliation":"同济大学 设计创新学院","sort_order":6},{"forum_id":4052,"role":"guest","name":"杨诚然","affiliation":"新加坡管理大学","sort_order":7}],"host":"","image_url":"/api/forums/4052/cover"},{"id":4071,"title":"The Third Brain-Computer Hybrid Intelligence: Toward Co-evolution of Biological and Machine Intelligence","topic":"Unclassified","kind":"专题论坛","chair":"赵莎、李发礼","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-516","description":"Brain-computer hybrid intelligence is an innovative concept that deeply integrates human intelligence and machine intelligence. It can not only leverage the advantages of computers in search, computation, storage, and optimization, but also compensate for human shortcomings in perception, reasoning, learning, and creativity. Through the complementarity of the respective advantages of humans and machines, the system's cognitive, perceptual, and behavioral capabilities are significantly enhanced. Is hybrid intelligence the future of artificial intelligence? From the current situation, the prospects are clear. Brain-computer hybrid intelligence can fully utilize the respective advantages of humans and machines, maximize the potential of both sides, and achieve synergistic enhancement.\n      Points for debate: 1. How to design a natural and efficient new paradigm for brain-computer interaction? 2. How to build neural encoding models for brain-computer interfaces? 3. How to achieve efficient control of complex objects based on brain-computer interfaces? 4. How to achieve enhancement of brain perception and senses based on brain-computer interfaces?","is_demo":0,"cover_url":"/api/forums/4071/cover?size=thumb","cover_alt":"The Third Brain-Computer Hybrid Intelligence: Toward Co-evolution of Biological and Machine Intelligence","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4071","source_checked_at":"2026-10-30","agenda_json":"[[\"Magnetic Resonance Brain Imaging Genetics and Individualized Neuromodulation for Psychiatric Disorders\",\"陈华富\"],[\"Research and Applications of Transcranial Focused Ultrasound\",\"杨健\"],[\"Brain Imaging-Pathology-Gene Fusion Analysis\",\"李志成\"],[\"Research on Artificial Intelligence Problems in Computational Cardiology\",\"骆功宁\"],[\"Neuroscience-Inspired Brain-Like Computing\",\"杨双鸣\"]]","code":"CNCC2026-4071","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-516","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4071","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第三届脑机混合智能：迈向生物智能与机器智能协同进化","original_topic":"未分类","original_description":"脑机混合智能是一种深度融合人类智能和机器智能的创新理念。它不仅能够发挥计算机在搜索、计算、存储和优化等方面的优势，还能弥补人类在感知、推理、学习和创造等方面的不足。通过人类和机器各自优势的互补，系统在认知、感知和行为能力方面得到了显著增强。混合智能是人工智能的未来吗？从目前的情况来看，前景是明确的。脑机混合智能能够充分利用人类和机器各自的优势，最大化地发挥双方的潜力，实现协同增效。\n      思辨点：1．如何设计自然高效的脑机交互新范式？2．如何构建面向脑机接口的神经编码模型？3．如何基于脑机接口实现复杂对象的高效控制？4．如何基于脑机接口实现大脑感知觉增强？","people":[{"forum_id":4071,"role":"chair","name":"赵莎","affiliation":"浙江大学","sort_order":0},{"forum_id":4071,"role":"guest","name":"陈华富","affiliation":"电子科技大学","sort_order":0},{"forum_id":4071,"role":"co_chair","name":"李发礼","affiliation":"电子科技大学","sort_order":1},{"forum_id":4071,"role":"guest","name":"杨健","affiliation":"北京理工大学","sort_order":1},{"forum_id":4071,"role":"guest","name":"李志成","affiliation":"中国科学院深圳先进技术研究院","sort_order":2},{"forum_id":4071,"role":"guest","name":"骆功宁","affiliation":"哈尔滨工业大学","sort_order":3},{"forum_id":4071,"role":"guest","name":"杨双鸣","affiliation":"天津大学","sort_order":4}],"host":"","image_url":"/api/forums/4071/cover"},{"id":4092,"title":"7th Neuromorphic Computing Forum — World Models: Algorithms and Computing Power","topic":"Vision, Multimodality and World Models","kind":"专题论坛","chair":"黄铁军","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城2号馆-207","description":"This forum is a featured special forum of CNCC2026 and has been held for seven consecutive editions. This year's brain-inspired computing forum focuses on the core direction of \"World Models: Algorithms and Computing Power.\" At present, general artificial intelligence is iterating rapidly, and world models have become a key foundation for achieving intelligent perception, reasoning, and prediction; however, they face challenges such as large computational demands and high real-time requirements, and their deployment requires breakthroughs in bottlenecks including architectural innovation, software-hardware co-design, algorithm optimization, and computing power adaptation. This forum attempts to connect model innovation with computing power breakthroughs and to promote the establishment of an efficient computing system for world models.\nThis forum has invited five experts in frontier exploration of world models and research on neuromorphic computing (in no particular order): Zhu Jun, Zhu Zheng, and Huang Yan focus on core algorithm innovation for world models, brain-inspired learning paradigms, model lightweighting, and optimization of general capabilities; Yang Yuchao and Xu Xiaoxin have deep expertise in intelligent computing power systems and conduct in-depth discussions on key technologies such as novel computing power architectures, compute-storage collaboration, and computing power adaptation and optimization.\nThe forum Panel aims to analyze the pain points and technology trends in world model R&D, explore feasible paths for brain-inspired computing to empower world models, and support the high-quality development of general artificial intelligence technology and its ecosystem.","is_demo":0,"cover_url":"/api/forums/4092/cover?size=thumb","cover_alt":"7th Neuromorphic Computing Forum — World Models: Algorithms and Computing Power","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4092","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Remarks\",\"黄铁军\"],[\"General World Models: First Principles and Roadmap\",\"朱军\"],[\"World Models: Unlocking Unlimited Physical AGI Productivity\",\"朱政\"],[\"A Unified Framework for Embodied World Models\",\"黄岩\"],[\"Memristor-Based Neuromorphic Dynamics Chips\",\"杨玉超\"],[\"Energy-Efficient In-Memory Computing Architectures and Training Strategies for Sparse Cognition\",\"许晓欣\"],[\"Panel Discussion\",\"赵地\"]]","code":"CNCC2026-4092","forum_type":"专题论坛","format":null,"theme_code":"视觉、多模态与世界模型","theme_name":"视觉、多模态与世界模型","venue":"西博城2号馆","room":"西博城2号馆-207","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":7,"forum_id":"CNCC2026-4092","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第七届神经形态计算论坛——世界模型：算法与算力","original_topic":"视觉、多模态与世界模型","original_description":"本论坛是 CNCC2026 特色专题论坛，已连续举办七届，本届脑启发计算论坛聚焦 “世界模型：算法与算力” 核心方向。当前通用人工智能快速迭代，世界模型成为实现智能感知、推理与预测的关键底座，但存在计算量大、实时性要求高等挑战，其落地需要突破架构创新、软硬协同、算法优化、算力适配等瓶颈难题，本论坛尝试连通模型创新与算力突破，推动建立世界模型高效计算体系。\n本论坛邀请了世界模型前沿探索和神经形态计算研究的五位专家（排名不分先后）：朱军、朱政、黄岩三位专家聚焦世界模型核心算法创新、脑启发学习范式、模型轻量化与通用能力优化；杨玉超、许晓欣两位专家深耕智能算力体系，围绕新型算力架构、存算协同、算力适配优化等关键技术展开深入研讨。\n论坛Panel旨在剖析世界模型研发痛点与技术趋势，探索脑启发计算赋能世界模型的可行路径，助力通用人工智能技术与生态的高质量发展。","people":[{"forum_id":4092,"role":"chair","name":"黄铁军","affiliation":"北京大学","sort_order":0},{"forum_id":4092,"role":"guest","name":"朱军","affiliation":"清华大学","sort_order":0},{"forum_id":4092,"role":"guest","name":"朱政","affiliation":"北京极佳视界科技有限公司","sort_order":1},{"forum_id":4092,"role":"guest","name":"黄岩","affiliation":"中国科学院自动化研究所","sort_order":2},{"forum_id":4092,"role":"guest","name":"杨玉超","affiliation":"北京大学","sort_order":3},{"forum_id":4092,"role":"guest","name":"许晓欣","affiliation":"中科院微电子所","sort_order":4},{"forum_id":4092,"role":"guest","name":"赵地","affiliation":"中国科学院计算技术研究所","sort_order":5}],"host":"","image_url":"/api/forums/4092/cover"},{"id":4102,"title":"Data Systems in the Intelligent Era","topic":"Data Science and Data Spaces","kind":"专题论坛","chair":"张岩峰","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城2号馆-201","description":"Against the backdrop of the rapid development of artificial intelligence and large model technologies, data systems are undergoing a profound transformation from traditional architectures to intelligent paradigms. How to achieve efficient and intelligent data storage and management in high-dimensional, multi-source, and heterogeneous data environments has become a frontier topic of common concern to academia and industry. This forum brings together multiple authoritative scholars in the database field to conduct in-depth discussion around core directions such as the construction and collaboration of data agents, high-dimensional vector retrieval for large models, AI-driven database autonomy and optimization, and new data system architectures. The forum aims to sort out the key breakthroughs in database technology in the intelligent era, explore innovative paths for the deep integration of data management and artificial intelligence, and provide a platform for intellectual exchange and academic communication for independent innovation and industrial upgrading in China's data systems field.","is_demo":0,"cover_url":"/api/forums/4102/cover?size=thumb","cover_alt":"Data Systems in the Intelligent Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4102","source_checked_at":"2026-10-30","agenda_json":"[[\"Intelligent Analysis Technologies for Large-Scale Structured Data\",\"袁野\"],[\"The Library of Large Models: Opportunities and Challenges of Vector Data Engines\",\"童咏昕\"],[\"Data-Driven Intelligent Self-Evolution\",\"周文超\"],[\"Agent-Driven Autonomous Data Science Systems: Progress and Prospects\",\"范举\"],[\"AlayaDB.AI: The Digital Foundation in the Era of Large Models\",\"唐博\"]]","code":"CNCC2026-4102","forum_type":"专题论坛","format":null,"theme_code":"数据科学与数据空间","theme_name":"数据科学与数据空间","venue":"西博城2号馆","room":"西博城2号馆-201","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4102","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智能时代的数据系统","original_topic":"数据科学与数据空间","original_description":"在人工智能与大模型技术迅猛发展的背景下，数据系统正经历从传统架构向智能化范式的深刻变革。如何在高维、多源、异构的数据环境中实现高效、智能的数据存储与管理，已成为学术界与产业界共同关注的前沿议题。本论坛汇聚数据库领域多位权威学者，围绕数据智能体的构建与协作、面向大模型的高维向量检索、AI驱动的数据库自治与优化、新型数据系统架构等核心方向展开深入研讨。论坛旨在梳理智能时代数据库技术的关键突破口，探索数据管理与人工智能深度融合的创新路径，为我国数据系统领域的自主创新与产业升级提供思想碰撞与学术交流的平台。","people":[{"forum_id":4102,"role":"guest","name":"袁野","affiliation":"北京理工大学","sort_order":0},{"forum_id":4102,"role":"chair","name":"张岩峰","affiliation":"东北大学","sort_order":0},{"forum_id":4102,"role":"guest","name":"童咏昕","affiliation":"北京航空航天大学","sort_order":1},{"forum_id":4102,"role":"guest","name":"周文超","affiliation":"阿里巴巴集团","sort_order":2},{"forum_id":4102,"role":"guest","name":"范举","affiliation":"中国人民大学","sort_order":3},{"forum_id":4102,"role":"guest","name":"唐博","affiliation":"南方科技大学数据库组","sort_order":4}],"host":"","image_url":"/api/forums/4102/cover"},{"id":4107,"title":"The First Space Supercomputing Forum","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"龚春叶","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城2号馆-202","description":"As Moore's Law fails and information processing capability approaches physical limits, space supercomputing will become the information infrastructure for humanity's outer space exploration and provide support for humanity's space migration. Space supercomputing faces challenges in system architecture technology, megawatt-level space power generation and energy storage technology, heat dissipation technology, communication technology, reliability technology, operations and maintenance technology, reusable launch technology, low-cost technology, and sustainable operations. Space supercomputing plays important application value in supporting space base operations and space-based computing, and in driving space power for space use, reusable launch, the development of space information technology, and the development of the low-altitude economy, and has excellent industrialization prospects. This forum focuses on the key technologies and application prospects of space supercomputing, aiming to build a platform for in-depth exchange in fields such as space supercomputing, space computing power, and computing power in space at home and abroad, and to promote collaborative research across industry, academia, and research.","is_demo":0,"cover_url":"/api/forums/4107/cover?size=thumb","cover_alt":"The First Space Supercomputing Forum","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4107","source_checked_at":"2026-10-30","agenda_json":"[[\"Space Supercomputing: The Inevitable Path for Humanity's Expansion into Space\",\"龚春叶\"],[\"Space-Ground Cross-Domain Digital-Intelligent Fusion Facilities and Future Applications\",\"孟祥飞\"],[\"Exploration of Semantic Computing Applications in Satellite-Ground Systems\",\"李聪\"],[\"Space-Ground Integrated Computing: Development Vision, Technology Evolution, and Implementation Path of Space Computing Centers\",\"刘波\"],[\"Counterintuitive Space Supercomputing and High-Compute Space Chip Design\",\"韩银和\"]]","code":"CNCC2026-4107","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城2号馆","room":"西博城2号馆-202","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4107","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"首届太空超级计算论坛","original_topic":"新型计算与网络架构","original_description":"随着摩尔定律失效，信息处理能力逼近物理极限，太空超级计算将成为人类进行外太空探索的信息基础设施，为人类文明的太空移民提供支撑。太空超级计算存在体系架构技术、兆瓦级太空发电及储能技术、散热技术、通信技术、可靠性技术、运维技术、可重复使用运载技术、低成本技术和可持续运营等挑战。太空超级计算在支撑太空基地运营和天数天算，牵引天电天用、可重复使用运载、太空信息技术发展和低空经济发展等方面发挥重要应用价值，并具有极好的产业化前景。本论坛聚焦太空超级计算关键技术及应用前景，旨在搭建国内外太空超级计算、太空算力、算力上天等领域深度交流平台，推动产学研协同攻关。","people":[{"forum_id":4107,"role":"chair","name":"龚春叶","affiliation":"国家超级计算天津中心","sort_order":0},{"forum_id":4107,"role":"guest","name":"孟祥飞","affiliation":"国家超级计算天津中心","sort_order":0},{"forum_id":4107,"role":"guest","name":"李聪","affiliation":"西安空间无线电技术研究所","sort_order":1},{"forum_id":4107,"role":"guest","name":"刘波","affiliation":"北京控制工程研究所","sort_order":2},{"forum_id":4107,"role":"guest","name":"韩银和","affiliation":"中国科学院计算技术研究所","sort_order":3}],"host":"","image_url":"/api/forums/4107/cover"},{"id":4108,"title":"How Can Data Engineering Become the Intelligent Foundation in the Era of Large Models?","topic":"Data Science and Data Spaces","kind":"专题论坛","chair":"丁小欧、王宏志","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-510","description":"As large models rapidly move toward continuous training, complex reasoning, retrieval augmentation, agent collaboration, and online evolution, data is upgrading from passively consumed \"fuel\" to \"soil\" that continuously nurtures model capabilities. However, traditional data engineering has long centered on storage, cleaning, integration, and offline governance. Faced with new problems such as multi-source heterogeneity, multimodality, real-time incrementality, high-value filtering, training-evaluation contamination, and synthetic data degradation, it can no longer fully support the continuous improvement of large model capabilities. Data engineering in the era of large models is shifting from management oriented toward data scale to management oriented toward the evolution of model capabilities.\n\nWith the theme \"How Can Data Engineering Become the Intelligent Foundation in the Era of Large Models?\", this forum focuses on the renewal and breakthrough of data engineering under the wave of large models, and discusses key frontier directions throughout the full lifecycle of large models, including data preparation, quality assessment, semantic governance, value measurement, and RAG/Agent data management. The forum will focus on deliberating three types of questions: first, the question of technical routes, namely how to support the upgrading of data engineering in the era of large models by building larger-scale data pipelines and data ecosystems oriented toward intelligent evolution; second, the question of value measurement, namely how to move from traditional scale and quality metrics to evaluating the marginal contribution to model capability improvement and establish a data contribution measurement system oriented toward large models; and third, the question of co-evolution, namely how to build a data engineering system in which data, models, and application environments evolve together for dynamic scenarios such as RAG, Agents, and online feedback, so as to support real-time knowledge updates and continuous capability improvement while avoiding data contamination, feedback bias, and capability degradation.\n\nThis forum has invited experts and scholars from top universities, research teams, and leading industry scenarios to engage in cross-domain dialogue, aiming to break the technical narrative of mere \"model involution\" and guide academia and industry to re-examine the strategic value of innovation on the data side. The forum will jointly explore how data engineering can be upgraded into a core intelligent foundation supporting continuous training, trustworthy alignment, knowledge updating, agent operation, and industry deployment of large models, providing new research perspectives, technical paths, and industry inspiration for the sustainable evolution of intelligent systems in the era of large models.","is_demo":0,"cover_url":"/api/forums/4108/cover?size=thumb","cover_alt":"How Can Data Engineering Become the Intelligent Foundation in the Era of Large Models?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4108","source_checked_at":"2026-10-30","agenda_json":"[[\"Industry Practices in Data Governance in the AI Era\",\"欧阳森山\"],[\"Architecture and Applications of Industrial Time-Series Data Intelligence Bases\",\"王建民\"],[\"Reflections and Practice on AI-Empowered Aviation Manufacturing\",\"牟文平\"],[\"Multimodal Data Compression and Retrieval\",\"杨晓春\"],[\"Breaking Data Barriers: From Large-Scale Data Synthesis to Domain-Specific Intelligent Analysis\",\"高云君\"],[\"From Large Models to Agents: The Changing and Unchanging Nature of Hallucinations\",\"冯骁骋\"]]","code":"CNCC2026-4108","forum_type":"专题论坛","format":null,"theme_code":"数据科学与数据空间","theme_name":"数据科学与数据空间","venue":"西博城5号馆","room":"西博城5号馆-510","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4108","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"数据工程，如何成为大模型时代的智能基座？","original_topic":"数据科学与数据空间","original_description":"在大模型快速迈向持续训练、复杂推理、检索增强、智能体协作与在线演化的今天，数据正在从被动消耗的“燃料”，升级为持续孕育模型能力的“土壤”。然而，传统数据工程长期围绕存储、清洗、集成与离线治理展开，面对多源异构、多模态、实时增量、高价值筛选、训练评测污染与合成数据退化等新问题，已难以充分支撑大模型能力的持续提升。大模型时代的数据工程，正在从面向数据规模的管理，走向面向模型能力演化的管理。\n\n本论坛以“数据工程，如何成为大模型时代的智能基座？”为主题，聚焦大模型浪潮下数据工程的更新与突围，围绕大模型全生命周期中的数据准备、质量评估、语义治理、价值度量、RAG/Agent 数据管理等关键前沿方向展开讨论。论坛将重点思辨三类问题：一是技术路线之问，即如何通过构建更大规模的数据管道与面向智能演化的数据生态系统，支撑大模型时代的数据工程升级；二是价值度量之问，即如何从传统规模和质量指标，走向对模型能力提升的边际贡献评估，建立面向大模型的数据贡献度量体系；三是协同演化之问，即如何面向 RAG、Agent 与在线反馈等动态场景，构建数据、模型与应用环境协同演进的数据工程体系，既支撑知识实时更新和能力持续提升，又避免数据污染、反馈偏置与能力退化。\n\n本论坛已邀请来自顶尖高校、科研团队与头部产业场景的专家学者展开跨领域对话，旨在打破单纯“模型内卷”的技术叙事，引导学术界与产业界重新审视数据侧创新的战略价值。论坛将共同探讨数据工程如何升级为支撑大模型持续训练、可信对齐、知识更新、智能体运行与行业落地的核心智能基座，为大模型时代智能系统的可持续演化提供新的研究视角、技术路径与产业启发。","people":[{"forum_id":4108,"role":"chair","name":"丁小欧","affiliation":"哈尔滨工业大学计算学部","sort_order":0},{"forum_id":4108,"role":"guest","name":"欧阳森山","affiliation":"航空工业成都飞机工业（集团）有限责任公司","sort_order":0},{"forum_id":4108,"role":"guest","name":"王建民","affiliation":"清华大学","sort_order":1},{"forum_id":4108,"role":"co_chair","name":"王宏志","affiliation":"哈尔滨工业大学","sort_order":1},{"forum_id":4108,"role":"guest","name":"牟文平","affiliation":"成都飞机工业集团有限责任公司","sort_order":2},{"forum_id":4108,"role":"guest","name":"杨晓春","affiliation":"东北大学","sort_order":3},{"forum_id":4108,"role":"guest","name":"高云君","affiliation":"浙江大学","sort_order":4},{"forum_id":4108,"role":"guest","name":"冯骁骋","affiliation":"哈尔滨工业大学","sort_order":5}],"host":"","image_url":"/api/forums/4108/cover"},{"id":4112,"title":"Breaking the Compute Bottleneck: Building and Evolving the Operator Ecosystem for Domestic AI Chips","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"朱光辉、冯思远","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-402","description":"The domestic AI chip industry is developing rapidly, and the operator development system has become a key hub connecting underlying chip capabilities with upper-layer model applications. However, in practical applications, problems such as long operator adaptation cycles, high development thresholds, and fragmented toolchains are common, making it difficult to fully unleash hardware capabilities. Compared with the continuous iteration of chips themselves, the construction of the operator development ecosystem is relatively weak and has become an important bottleneck for the large-scale application of domestic computing power. This forum focuses on the \"operator development ecosystem,\" a key supporting layer above domestic AI chips, and starts from the role of operators in large model training and inference to conduct in-depth deliberation. The deliberation topics include:\n1) Where are the key bottlenecks through which the operator ecosystem constrains the performance release and large-scale application of AI chips?\n2) How can AI chip vendors, open-source communities, and other parties collaborate to jointly build a domestic operator ecosystem and achieve continuous evolution?\n3) In the AI Agents era, what new opportunities and paradigm innovations will the domestic operator development ecosystem bring?","is_demo":0,"cover_url":"/api/forums/4112/cover?size=thumb","cover_alt":"Breaking the Compute Bottleneck: Building and Evolving the Operator Ecosystem for Domestic AI Chips","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4112","source_checked_at":"2026-10-30","agenda_json":"[[\"From Programming to Search: TileLang-Driven Automatic Operator Generation\",\"杨智\"],[\"FlagMega: A Megakernel Compiler for the AI Agent Era\",\"郭晖\"],[\"Implementing Frontier LLM Technologies on Muxi GPUs\",\"董兆华\"],[\"\\\"Breakthrough, Symbiosis, Leadership\\\": How CANN Builds a New Ecosystem for Domestic Operators\",\"黄金华\"]]","code":"CNCC2026-4112","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-402","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4112","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"破局算力瓶颈：国产AI芯片算子生态的构建与演进","original_topic":"新型计算与网络架构","original_description":"国产AI芯片产业快速发展，算子开发体系已成为连接底层芯片能力与上层模型应用的关键枢纽。但实际应用中普遍面临算子适配周期长、开发门槛高、工具链碎片化等问题，导致硬件能力难以充分释放。相比芯片本身的持续迭代，算子开发生态建设相对薄弱，已成为国产算力规模化应用的重要瓶颈。本论坛聚焦“算子开发生态”这一国产AI芯片之上的关键支撑层，从算子在大模型训练与推理中的作用切入，展开深度思辨。思辨议题包括：\n1）算子生态制约AI芯片性能释放与规模化应用的关键瓶颈在哪里？\n2）AI芯片厂商与开源社区等各方如何协同共建国产算子生态并实现持续演进？\n3）在 Agent 时代国产算子开发生态会带来什么新的机遇与范式革新？","people":[{"forum_id":4112,"role":"chair","name":"朱光辉","affiliation":"南京大学","sort_order":0},{"forum_id":4112,"role":"guest","name":"杨智","affiliation":"北京大学","sort_order":0},{"forum_id":4112,"role":"co_chair","name":"冯思远","affiliation":"上海创智学院","sort_order":1},{"forum_id":4112,"role":"guest","name":"郭晖","affiliation":"北京智源人工智能研究院","sort_order":1},{"forum_id":4112,"role":"guest","name":"董兆华","affiliation":"沐曦集成电路（上海）有限公司","sort_order":2},{"forum_id":4112,"role":"guest","name":"黄金华","affiliation":"海思技术有限公司","sort_order":3},{"forum_id":4112,"role":"guest","name":"陈果","affiliation":"湖南大学","sort_order":4},{"forum_id":4112,"role":"guest","name":"陈振宇","affiliation":"南京大学","sort_order":5},{"forum_id":4112,"role":"guest","name":"王腾","affiliation":"中国科学技术大学苏州高等研究院","sort_order":6}],"host":"","image_url":"/api/forums/4112/cover"},{"id":4118,"title":"RF Computing: Exploring the Boundaries of Intelligence","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"王巍、郭秀珍","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-414","description":"From signal perception to spatial cognition, RF computing is moving from the processing of physical signals toward the reshaping of intelligent forms. RF computing refers to the process of transforming RF signals within the RF space to accomplish information processing and conversion, and its connotation is continuously expanding with the deep penetration of artificial intelligence.\n\nThis forum takes \"RF computing\" as its core theme and, building on the discussions of the previous two editions, further introduces frontier perspectives such as large models, embodied intelligence, and edge computing: theoretically, it deepens the scientific connotation and improves the theoretical system; technologically, it focuses on information metamaterials, intelligent reflecting surfaces, passive Internet of Things, and intelligent computing in the RF domain; and in application, it expands forward-looking scenarios such as the low-altitude economy, digital twins, smart cities, and AI-enabled communications.\n\nThe forum will invite renowned experts and scholars in the field to give invited talks, engage in in-depth exchanges with attending peers, build consensus on interdisciplinary innovation, and jointly explore the infinite possibilities of RF computing in the intelligent era.","is_demo":0,"cover_url":"/api/forums/4118/cover?size=thumb","cover_alt":"RF Computing: Exploring the Boundaries of Intelligence","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4118","source_checked_at":"2026-10-30","agenda_json":"[[\"New Progress in Integrated Sensing, Communication and Computing for Intelligent Metasurfaces\",\"程强\"],[\"RF Computing: Progress and Reflections\",\"何源\"],[\"Networked Spatial Embodied Intelligence\",\"鲁力\"],[\"Latest Progress in 6G-Oriented Mobile Passive IoT\",\"龚伟\"],[\"Backscatter Communication Technology for 6G Passive IoT\",\"于季弘\"]]","code":"CNCC2026-4118","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-414","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4118","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"射频计算：探索智能边界","original_topic":"新型计算与网络架构","original_description":"从信号感知到空间认知，射频计算正从对物理信号的处理走向对智能形态的重塑。射频计算是指在射频空间内对射频信号做变换以完成信息处理与转换的过程，其内涵正随人工智能的深度渗透而不断拓展。\n\n本届论坛以\"射频计算\"为核心主题，在前两届探讨的基础上，进一步引入大模型、具身智能、边缘计算等前沿视角：理论上深化科学内涵、完善理论体系，技术上聚焦信息超材料、智能反射面、无源物联网及射频域智能计算，应用上拓展低空经济、数字孪生、智慧城市、AI赋能通信等前瞻场景。\n\n论坛将邀请领域内著名专家学者作特邀报告，与参会同行深入交流，凝聚交叉创新共识，共同探索智能时代下射频计算的无限可能。","people":[{"forum_id":4118,"role":"chair","name":"王巍","affiliation":"武汉大学","sort_order":0},{"forum_id":4118,"role":"guest","name":"程强","affiliation":"东南大学","sort_order":0},{"forum_id":4118,"role":"co_chair","name":"郭秀珍","affiliation":"浙江大学","sort_order":1},{"forum_id":4118,"role":"guest","name":"何源","affiliation":"清华大学","sort_order":1},{"forum_id":4118,"role":"guest","name":"鲁力","affiliation":"电子科技大学","sort_order":2},{"forum_id":4118,"role":"guest","name":"龚伟","affiliation":"中国科学技术大学","sort_order":3},{"forum_id":4118,"role":"guest","name":"于季弘","affiliation":"北京理工大学","sort_order":4}],"host":"","image_url":"/api/forums/4118/cover"},{"id":4119,"title":"Educating the Future: Reshaping and Advancing the Role of University Teachers in the AI Era","topic":"General","kind":"专题论坛","chair":"周傲英、龚声蓉","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-513","description":"At present, the new generation of artificial intelligence technologies centered on large models and AI agents is accelerating with disruptive momentum, profoundly reshaping the knowledge boundaries, teaching systems, and engineering practice paradigms of computer science. To actively respond to the national strategic deployment of \"AI + Education\" and closely follow the core direction of digital transformation in higher education, this forum takes as its theme \"Educating the Future, Human-Machine Coexistence: Reshaping and Advancing the Role of University Teachers in the AI Era,\" and conducts in-depth discussion from multiple dimensions including teaching model innovation, knowledge system reconstruction, upgrading educational concepts, integration of research capabilities, and improvement of professional competence. It focuses on the role transformation of teachers under AI empowerment from \"knowledge transmitters\" to \"learning guides, technology masters, innovation cultivators, and value leaders,\" and systematically sorts out the core paths and practical methods for the transformation of university teachers.","is_demo":0,"cover_url":"/api/forums/4119/cover?size=thumb","cover_alt":"Educating the Future: Reshaping and Advancing the Role of University Teachers in the AI Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4119","source_checked_at":"2026-10-30","agenda_json":"[[\"Contemporary Artificial Intelligence Is the Hallmark of a New Human Civilization\",\"周傲英\"],[\"The Mission and Responsibility of University Teachers in the Age of Artificial Intelligence—Exploration and Practice at Northeastern University\",\"王兴伟\"],[\"Digital-Intelligent Teaching-Research Integration Driving the Enhancement of University Teachers' Research and Innovation Capabilities\",\"郝天永\"],[\"AI Empowerment: Building a \\\"Golden Textbook\\\" Ecosystem for Teaching-Friendly and Learning-Friendly Vocational Education\",\"武春岭\"],[\"A New Creative Engine in the AI Era—How YunGongKaiWu Empowers Young Teachers\",\"李贝\"]]","code":"CNCC2026-4119","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城5号馆","room":"西博城5号馆-513","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4119","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智育未来：AI 时代高校教师角色重塑与进阶","original_topic":"综合","original_description":"当前，以大模型、智能体为核心的新一代人工智能技术正以颠覆性态势加速演进，深刻重塑着计算机学科的知识边界、教学体系与工程实践范式。为积极响应国家 “人工智能 + 教育” 的战略部署，紧扣高等教育数字化转型核心方向，本论坛特以 “智育未来，人机共生：AI 时代高校教师角色重塑与进阶” 为主题，从教学模式创新、知识体系重构、育人理念升级、科研能力融合、职业素养提升等多维度展开深度探讨，聚焦 AI 赋能下教师从 “知识传授者” 向 “学习引导者、技术驾驭者、创新培育者、价值引领者” 的角色转变，系统梳理高校教师转型的核心路径与实践方法。","people":[{"forum_id":4119,"role":"chair","name":"周傲英","affiliation":"华东师范大学","sort_order":0},{"forum_id":4119,"role":"guest","name":"王兴伟","affiliation":"东北大学","sort_order":0},{"forum_id":4119,"role":"co_chair","name":"龚声蓉","affiliation":"苏州工学院","sort_order":1},{"forum_id":4119,"role":"guest","name":"郝天永","affiliation":"华南师范大学","sort_order":1},{"forum_id":4119,"role":"guest","name":"武春岭","affiliation":"重庆电子科技职业大学","sort_order":2},{"forum_id":4119,"role":"guest","name":"李贝","affiliation":"阿里巴巴","sort_order":3},{"forum_id":4119,"role":"guest","name":"车翔玖","affiliation":"吉林大学计算机科学与技术学院","sort_order":4},{"forum_id":4119,"role":"guest","name":"王奔","affiliation":"杭州师范大学","sort_order":5},{"forum_id":4119,"role":"guest","name":"李航","affiliation":"沈阳师范大学","sort_order":6}],"host":"","image_url":"/api/forums/4119/cover"},{"id":4123,"title":"Large Models Enter the Production Line: Breaking Through the Barriers of Cloud-Edge Collaboration from Industrial Large Models \"Running\" to \"Working Well\"","topic":"AI for Industry","kind":"专题论坛","chair":"张永敏、郭泰彪","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-410","description":"Large model technology is accelerating from the \"chat room\" to the \"production line,\" showing huge potential in scenarios such as industrial quality inspection, equipment operation and maintenance, production scheduling, and safety monitoring. However, most current cases still remain at the Demo stage of \"running.\" Can we truly achieve seamless, efficient, and deployable cloud-edge collaborative industrial large models? In real deployment of industrial large models, what are the real technical bottlenecks faced by cloud-edge collaboration? And where are the pain points in the implementation process? Let us jointly analyze the root causes in depth and provide new ideas for breaking through the dilemma.\nDebate point 1: Which industrial application scenarios require cloud-edge collaborative computing for large models?\nDebate point 2: Is the core technical bottleneck of large-model cloud-edge collaboration the \"data wall,\" the \"model wall,\" the \"communication wall,\" or the \"computing power wall\"?\nDebate point 3: From technical feasibility to industry willingness to use, what are the key issues urgently to be solved for production-line large models? How can we break through?","is_demo":0,"cover_url":"/api/forums/4123/cover?size=thumb","cover_alt":"Large Models Enter the Production Line: Breaking Through the Barriers of Cloud-Edge Collaboration from Industrial Large Models \"Running\" to \"Working Well\"","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4123","source_checked_at":"2026-10-30","agenda_json":"[[\"Cloud-Edge Collaborative Metallurgical Vision: Multimodal Perception and Trustworthy Intelligence under Extreme Conditions\",\"蒋朝辉\"],[\"Multimodal Industrial Anomaly Detection: Unified Architecture and Applications\",\"贺诗波\"],[\"Zongheng Intelligence: Practice and Reflections on Cloud-Edge Collaboration and Large-Small Model Collaboration\",\"王田\"],[\"From \\\"Running Through\\\" to \\\"Continuously Creating Value\\\": Knowledge and Agent Architecture Design for Scaled Industrial Agent Deployment\",\"武立忠\"],[\"From \\\"Seeing\\\" to \\\"Understanding\\\": Large-Model-Driven Intelligent Recognition and Rating of TEM/SEM Microstructures\",\"朱焱麟\"]]","code":"CNCC2026-4123","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"西博城4号馆","room":"西博城4号馆-410","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4123","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"大模型进产线：云边协同从工业大模型“能跑通”到“真好用”的破壁之路","original_topic":"AI+行业","original_description":"大模型技术正加速从“聊天室”走向“生产线”，在工业质检、设备运维、生产调度、安全监控等场景中展现出了巨大潜力。然而，当前绝大多数案例仍停留在“能跑通”的Demo阶段。我们能否真正实现无感、高效、可落地的云边协同工业大模型？在工业大模型真实部署中，云边协同面临着哪些真正的技术瓶颈？其落地过程中的痛点又在哪里？让我们共同深度剖析下其根本原因，为其破局提供新思路。\n思辨点1: 哪些工业应用场景的大模型需要云边协同计算？\n思辨点2: 大模型云边协同的核心技术瓶颈是“数据墙”、“模型墙”、“通信墙”还是“算力墙”？\n思辨点3: 从技术可行到工业愿意用，产线大模型的亟待解决的关键问题是什么？如何破局？","people":[{"forum_id":4123,"role":"chair","name":"张永敏","affiliation":"中南大学","sort_order":0},{"forum_id":4123,"role":"guest","name":"蒋朝辉","affiliation":"中南大学","sort_order":0},{"forum_id":4123,"role":"co_chair","name":"郭泰彪","affiliation":"云边智算科技有限公司","sort_order":1},{"forum_id":4123,"role":"guest","name":"贺诗波","affiliation":"浙江大学","sort_order":1},{"forum_id":4123,"role":"guest","name":"周欢","affiliation":"西北工业大学","sort_order":2},{"forum_id":4123,"role":"guest","name":"王田","affiliation":"北京师范大学","sort_order":3},{"forum_id":4123,"role":"guest","name":"武立忠","affiliation":"深信服科技股份有限公司","sort_order":4},{"forum_id":4123,"role":"guest","name":"朱焱麟","affiliation":"成都先进金属材料产业技术研究院股份有限公司","sort_order":5}],"host":"","image_url":"/api/forums/4123/cover"},{"id":4125,"title":"AI + Ocean: Mechanism Analysis and Scientific Discovery of Air-Sea Coupled Large Models","topic":"AI4Science","kind":"专题论坛","chair":"周小伟、李春山","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城2号馆-203","description":"Pangu and other ocean forecast large models are driving AI from a numerical model accelerator to the main body of ocean meteorological prediction, learning the relationships among atmospheric circulation, sea temperature and currents, and air-sea fluxes. The key to air-sea large models is not only jointly predicting multiple types of elements, but establishing computable, constrainable, and interpretable coupling channels within the model, and driving the model from variable correlation toward mechanism association through shared latent spaces, coupled data assimilation, and physical constraints. However, existing attention, attribution, and physical consistency checks still find it difficult to answer whether the model truly conforms to air-sea mechanisms. This forum focuses on the forecast accuracy, mechanism interpretation, cross-domain robustness, and operational application of air-sea large models.\nDebate:\n1. Where are the methodological spectrum and applicable boundaries of large model interpretability? 2. How can statistical dependence and physical mechanism representation in air-sea large models be distinguished? 3. For trustworthy prediction and operational application, how can an interpretable, constrainable, and verifiable air-sea coupled large model technology system be built?","is_demo":0,"cover_url":"/api/forums/4125/cover?size=thumb","cover_alt":"AI + Ocean: Mechanism Analysis and Scientific Discovery of Air-Sea Coupled Large Models","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4125","source_checked_at":"2026-10-30","agenda_json":"[[\"Intelligent Rapid Forecasting of Ocean Dynamic Disasters and Its Agents\",\"董昌明\"],[\"Knowledge-Data-Physics Integration: A New AI-Driven Approach to Complex Computable Ocean Systems\",\"聂婕\"],[\"Models and Methods for Short-Term Nowcasting in Ocean Regions\",\"李旭涛\"],[\"Development and Application of an AI Algorithm for Addressing Typhoon-Induced Gaps in Sea Surface Observations\",\"宫勋\"],[\"Jingwei: Intelligent Reconstruction and Cognition of Global Ocean Element Cycles\",\"卢彬\"],[\"TBD\",\"邓锦亮\"]]","code":"CNCC2026-4125","forum_type":"专题论坛","format":null,"theme_code":"AI4Science","theme_name":"AI4Science","venue":"西博城2号馆","room":"西博城2号馆-203","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4125","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI+海洋：海气耦合大模型的机理解析与科学发现","original_topic":"AI4Science","original_description":"盘古及其他海洋预报大模型，正推动AI由数值模式加速器转向海洋气象预测主体，学习大气环流、海温海流与海气通量关系。海气大模型的关键，不只是联合预测多类要素，而是在模型内部建立可计算、可约束、可解释的耦合通道，通过共享潜空间、耦合数据同化和物理约束，推动模型从变量相关走向机制关联。但现有注意力、归因和物理一致性检验，仍难回答模型是否真正符合海气机理。本论坛聚焦海气大模型的预报精度、机理解释、跨域稳健与业务应用。\n思辨：\n1.大模型可解释性的方法谱系与适用边界在哪？2.如何区分海气大模型中的统计依赖与物理机制表征？3.面向可信预测与业务应用，如何构建可解释、可约束、可验证的海气耦合大模型技术体系？","people":[{"forum_id":4125,"role":"chair","name":"周小伟","affiliation":"中国海洋大学","sort_order":0},{"forum_id":4125,"role":"guest","name":"董昌明","affiliation":"南京信息工程大学","sort_order":0},{"forum_id":4125,"role":"co_chair","name":"李春山","affiliation":"哈尔滨工业大学（威海）","sort_order":1},{"forum_id":4125,"role":"guest","name":"聂婕","affiliation":"中国海洋大学","sort_order":1},{"forum_id":4125,"role":"guest","name":"李旭涛","affiliation":"哈尔滨工业大学深圳研究生院","sort_order":2},{"forum_id":4125,"role":"guest","name":"宫勋","affiliation":"山东省计算中心","sort_order":3},{"forum_id":4125,"role":"guest","name":"卢彬","affiliation":"上海交通大学","sort_order":4},{"forum_id":4125,"role":"guest","name":"邓锦亮","affiliation":"北京航空航天大学","sort_order":5}],"host":"","image_url":"/api/forums/4125/cover"},{"id":4131,"title":"Integrating Industry and Research for New Opportunities: Forum on New Models of Industry-University-Research Collaboration in the AI Era","topic":"General","kind":"专题论坛","chair":"秦征、吴国斌","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"天府国际会议中心-104黄龙","description":"This CNCC2026 special forum on industry-university-research collaboration builds a cross-sector dialogue platform for corporate school-enterprise cooperation leaders, university industry-university-research management staff, and frontline research faculty, focusing on the new transformations in collaborative industry-university-research development under the AI wave, and jointly exploring new challenges and new opportunities for industry development.\n \nCurrently, AI technology is comprehensively reconstructing the entire chain of research breakthroughs, achievement transformation, and talent co-cultivation, and traditional industry-university-research collaboration models, value goals, and implementation paths face systemic change. The forum conducts in-depth discussion around three core topics: first, new industry-university-research collaboration paradigms suited to industrial upgrading and disciplinary development in the AI era, breaking down diverse implementation forms such as joint laboratories, targeted research, and industry-education communities; second, clarifying the core goals of industry-university-research integration and balancing the three core demands of technological breakthroughs, industrialization of achievements, and cultivation of interdisciplinary talent; third, sharing successful cases of school-enterprise collaboration in the AI era and distilling replicable and scalable collaboration operating mechanisms.","is_demo":0,"cover_url":"/api/forums/4131/cover?size=thumb","cover_alt":"Integrating Industry and Research for New Opportunities: Forum on New Models of Industry-University-Research Collaboration in the AI Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4131","source_checked_at":"2026-10-30","agenda_json":"[[\"Review of Industry-University-Research Development\",\"杨士强\"],[\"What Are the Goals of Industry-University-Research Collaboration from the Perspective of AI Chip Companies?\",\"杨言\"],[\"What Are the Goals of Industry-University-Research Collaboration from the Perspective of AI Application Companies?\",\"陈虹雨\"],[\"Exploration of Industry-University-Research Collaboration Models for Large Model Enterprises\",\"何芸\"],[\"Methods and Model Innovation of State-Owned Enterprise Research Institutes in Industry-University-Research Collaboration in the AI Era\",\"张健松\"]]","code":"CNCC2026-4131","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"天府国际会议中心","room":"天府国际会议中心-104黄龙","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4131","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智融产研·创领新机——AI时代产学研协同新模式发展论坛","original_topic":"综合","original_description":"本次CNCC2026产学研合作专项论坛，面向企业校企合作负责人、高校产学研管理工作者、一线科研教师搭建跨界对话平台，聚焦人工智能浪潮下产学研协同发展的全新变局，共探行业发展新挑战、新机遇 。\n \n当前AI技术全面重构科研攻关、成果转化、人才共育全链条，传统产学研合作模式、价值目标与落地路径面临系统性变革。论坛围绕三大核心议题深度研讨：一是AI时代适配产业升级、学科发展的新型产学研合作范式，拆解联合实验室、定向攻关、产教共同体等多元落地形态；二是厘清产学研融合核心目标，平衡技术突破、成果产业化、复合型人才培养三大核心诉求；三是分享AI时代校企协同成功案例，提炼可复制、可推广的合作运营机制。","people":[{"forum_id":4131,"role":"guest","name":"杨士强","affiliation":"清华大学","sort_order":0},{"forum_id":4131,"role":"chair","name":"秦征","affiliation":"OPPO广东移动通信有限公司","sort_order":0},{"forum_id":4131,"role":"co_chair","name":"吴国斌","affiliation":"滴滴出行","sort_order":1},{"forum_id":4131,"role":"guest","name":"张健松","affiliation":"中国电信股份有限公司","sort_order":1},{"forum_id":4131,"role":"guest","name":"陈虹雨","affiliation":"小米科技","sort_order":2},{"forum_id":4131,"role":"guest","name":"何芸","affiliation":"北京智谱华章科技有限公司","sort_order":3},{"forum_id":4131,"role":"guest","name":"杨言","affiliation":"摩尔线程智能科技（北京）股份有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4131/cover"},{"id":4143,"title":"Intelligence Made Tangible, Risk Unseen: From Point Protection to Hardware-Software Collaborative Governance for Embodied Intelligent Systems","topic":"Unclassified","kind":"专题论坛","chair":"张悦","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-407","description":"As embodied intelligence, humanoid robots, and AI Agents enter real-world scenarios, the IoT security boundary is expanding from single devices, protocols, and access control to the full chain of AI-driven perception, decision-making, and execution. Multimodal input, Agent invocation, and collaboration between large and small models may introduce environmental understanding bias, identity delegation, permission drift, and policy semantic bias, and further act on the physical world in the execution stage, affecting personnel, equipment, and production processes. However, the current security system is still mainly built separately around model security, application security, network security, and endpoint security, lacking a unified analysis framework that runs through the entire \"perception-decision-execution\" chain. The forum hopes to discuss the security of the AI control chain in embodied IoT, explore how to move from point protection to hardware-software collaborative governance, and build a new security system oriented to closed-loop control processes.","is_demo":0,"cover_url":"/api/forums/4143/cover?size=thumb","cover_alt":"Intelligence Made Tangible, Risk Unseen: From Point Protection to Hardware-Software Collaborative Governance for Embodied Intelligent Systems","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4143","source_checked_at":"2026-10-30","agenda_json":"[[\"Reflections on the Security of Embodied Intelligence (Unmanned Systems)\",\"孙利民\"],[\"Low-Altitude Intelligence: From Perceptual Reasoning to Swarm Embodied Security\",\"姚鑫杰\"],[\"Reliable Embodied Perception and Planning for Complex Tasks in 3D Environments\",\"冯拓\"],[\"Research on Security Infrastructure for the Agent Ecosystem\",\"姚苏\"]]","code":"CNCC2026-4143","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-407","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4143","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智能有形，风险无相：具身智能系统从单点防护到软硬件协同治理","original_topic":"未分类","original_description":"随着具身智能、人形机器人和 AI Agent 进入现实场景，物联网安全边界正从单一设备、协议和访问控制，扩展到由 AI 驱动的感知、决策与执行全链路。多模态输入、Agent 调用和大小模型协同可能带来环境理解偏差、身份委托、权限漂移和策略语义偏差，并在执行环节进一步作用于物理世界，影响人员、设备与生产过程。然而当前安全体系仍以模型安全、应用安全、网络安全和终端安全分别建设为主，缺乏贯穿“感知—决策—执行”全链路的统一分析框架。论坛希望围绕具身物联网中的AI控制链安全展开讨论，探索如何从单点防护走向软硬件协同治理，构建面向闭环控制过程的新型安全体系。","people":[{"forum_id":4143,"role":"chair","name":"张悦","affiliation":"山东大学","sort_order":0},{"forum_id":4143,"role":"guest","name":"孙利民","affiliation":"中科院信息工程研究所","sort_order":0},{"forum_id":4143,"role":"guest","name":"张翰林","affiliation":"青岛大学","sort_order":1},{"forum_id":4143,"role":"guest","name":"姚鑫杰","affiliation":"昆明理工大学","sort_order":2},{"forum_id":4143,"role":"guest","name":"冯拓","affiliation":"浙江大学","sort_order":3},{"forum_id":4143,"role":"guest","name":"姚苏","affiliation":"清华大学","sort_order":4}],"host":"","image_url":"/api/forums/4143/cover"},{"id":4155,"title":"4th Homomorphic Encryption Computing Acceleration Technology Forum — Cryptography, Chips, and Systems","topic":"Unclassified","kind":"专题论坛","chair":"张明喆","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-507","description":"Homomorphic encryption enables computation directly on encrypted data and has already seen preliminary applications in fields such as cloud computing, healthcare, and fintech. However, encrypted computation systems based on homomorphic encryption still face certain cost constraints and cannot do without close collaboration across multiple fields including cryptography, algorithms, systems, architecture, and circuits. This homomorphic encryption computing acceleration technology forum will invite experts engaged in research in homomorphic encryption-related fields from home and abroad to discuss the core challenges faced in the process of moving homomorphic encryption technology from theory to practice, exchange the latest research progress, form innovative thinking and constructive opinions, and promote the further development of homomorphic encryption.","is_demo":0,"cover_url":"/api/forums/4155/cover?size=thumb","cover_alt":"4th Homomorphic Encryption Computing Acceleration Technology Forum — Cryptography, Chips, and Systems","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4155","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Report: How Far Are We from Practical Homomorphic Encryption Computing\",\"张明喆\"],[\"Research Progress on Fully Homomorphic Encryption Virtual Machines\",\"边松\"],[\"Privacy-Preserving Computing Acceleration Based on a Unified Hardware Architecture and Flexible Operator Mapping\",\"高鸣宇\"],[\"Fully Homomorphic Encryption Chips and Their Applications\",\"刘雷波\"],[\"Accelerating Homomorphic Encryption: A Seven-Year Journey\",\"Jung Ho Ahn\"],[\"Asynchronous Lagrange-Based Threshold FHE with Smaller Modulus Overhead\",\"Changmin Lee\"]]","code":"CNCC2026-4155","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-507","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4155","legacy_id":3,"lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第四届同态加密计算加速技术论坛——密码、芯片与系统","original_topic":"未分类","original_description":"同态加密可在密文数据上直接进行计算，已在云计算、医疗健康、金融科技等领域具有了初步的应用。然而，基于同态加密的密态计算体系仍然面临一定的成本制约，离不开密码学、算法、系统、体系结构、电路等多个领域间的紧密协作。本届同态加密计算加速技术论坛将邀请国内外从事同态加密相关领域研究的专家，就同态加密的技术从理论走向实践过程中面临的核心挑战进行讨论，并交流最新研究进展，形成创新性思考和建设性意见，推动同态加密的进一步发展。","people":[{"forum_id":4155,"role":"chair","name":"张明喆","affiliation":"蚂蚁技术研究院","sort_order":0},{"forum_id":4155,"role":"guest","name":"边松","affiliation":"北京航空航天大学","sort_order":0},{"forum_id":4155,"role":"guest","name":"高鸣宇","affiliation":"清华大学","sort_order":1},{"forum_id":4155,"role":"guest","name":"刘雷波","affiliation":"清华大学","sort_order":2},{"forum_id":4155,"role":"guest","name":"Jung Ho Ahn","affiliation":"Seoul National University","sort_order":3},{"forum_id":4155,"role":"guest","name":"Changmin Lee","affiliation":"Korea University","sort_order":4}],"host":"","image_url":"/api/forums/4155/cover"},{"id":4157,"title":"Output Carriers and Evaluation Systems under the Research Automation Paradigm","topic":"General","kind":"专题论坛","chair":"刘勇、张忠宝","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-506","description":"Currently, research automation technology is accelerating its penetration into the entire chain of academic research, driving a profound transformation of the research paradigm from traditional human-driven to human-machine collaboration, and the \"acceleration effect\" of academic research is becoming increasingly prominent. The deep follow-up questions that ensue have also attracted widespread attention in academia: Are automation tools the core driving force that liberates researchers from repetitive labor and fully unleashes creativity, or a potential risk that blurs the boundaries of originality of academic outputs, dissolves the value of peer review, and undermines the credibility of academic evaluation systems?\nWith the evolution of the research automation paradigm, academic output carriers are breaking through the solidified form of traditional papers, and the dimensions for determining output value and the rules for attributing contributions have undergone profound changes, further intensifying the contradiction between existing evaluation systems and research practice. The traditional evaluation framework centered on papers and based on peer review is encountering practical challenges when facing automatically generated research outputs, such as increased difficulty in identifying originality, difficulty in quantifying process value, and insufficient adaptability of talent evaluation standards.\nThis YOCSEF viewpoint forum takes \"Output Carriers and Evaluation Systems under the Research Automation Paradigm\" as its theme and focuses on three core debate topics: (1) Output carriers — papers, patents, data, models, etc. — which is the true representative of \"research outputs\"? (2) Evaluation systems — when research productivity grows explosively, how should \"academic value\" be redefined? (3) Under such output forms, where should human capabilities return to? The forum invites experts from various fields and young research practitioners to engage in in-depth speculation, aiming to clarify the deep impact of research automation on the academic ecosystem through the collision of multi-position viewpoints, explore evaluation systems adapted to new research paradigms, and provide forward-looking thinking for building an efficient and fair academic evaluation ecosystem.","is_demo":0,"cover_url":"/api/forums/4157/cover?size=thumb","cover_alt":"Output Carriers and Evaluation Systems under the Research Automation Paradigm","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4157","source_checked_at":"2026-10-30","agenda_json":"[]","code":"CNCC2026-4157","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城5号馆","room":"西博城5号馆-506","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":0,"forum_id":"CNCC2026-4157","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"科研自动化范式下的成果载体与评价体系","original_topic":"综合","original_description":"当前，科研自动化技术加速渗透学术研究全链条，推动科研范式由传统人力驱动向人机协同模式深刻转型，学术研究的 “加速效应” 日益凸显。随之而来的深层追问也引发学界广泛关注：自动化工具究竟是将科研人员从重复性劳动中解放、充分释放创造力的核心动能，还是会模糊学术成果原创性边界、消解同行评议价值、冲击学术评价体系公信力的潜在风险？\n伴随科研自动化范式演进，学术成果载体突破传统论文固化形态，成果价值判定维度、贡献归属规则均发生深刻变化，进一步加剧了现有评价体系与科研实践脱节的矛盾。传统以论文为核心、基于同行评议的评价框架，在面对自动化生成的科研产出时，正遭遇原创性甄别难度提升、过程性价值难以量化、人才评价标准适配性不足等现实挑战。\n本次 YOCSEF 观点论坛以 “科研自动化范式下的成果载体与评价体系” 为主题，聚焦三项核心辩题：（1）成果载体 —— 论文、专利、数据、模型等，谁才是‘科研成果’的真正代表？（2）评价体系 —— 当科研生产力爆发式增长，‘学术价值’该如何重新定义？（3）在此类成果形态下，人的能力应该回归到何处？论坛邀请各领域专家与青年科研从业者展开深度思辨，旨在通过多立场观点碰撞，厘清科研自动化对学术生态的深层影响，探索适配新型科研范式的评价体系，为构建高效公正的学术评价生态提供前瞻性思考。","people":[{"forum_id":4157,"role":"chair","name":"刘勇","affiliation":"中国人民大学","sort_order":0},{"forum_id":4157,"role":"guest","name":"文继荣","affiliation":"中国人民大学","sort_order":0},{"forum_id":4157,"role":"co_chair","name":"张忠宝","affiliation":"北京邮电大学","sort_order":1},{"forum_id":4157,"role":"guest","name":"何向南","affiliation":"中国科学技术大学","sort_order":1},{"forum_id":4157,"role":"guest","name":"陈思衡","affiliation":"上海交通大学","sort_order":2},{"forum_id":4157,"role":"guest","name":"雷印杰","affiliation":"四川大学","sort_order":3},{"forum_id":4157,"role":"guest","name":"张宁豫","affiliation":"浙江大学","sort_order":4},{"forum_id":4157,"role":"guest","name":"许洪腾","affiliation":"中国人民大学","sort_order":5},{"forum_id":4157,"role":"guest","name":"陈科海","affiliation":"哈尔滨工业大学（深圳）","sort_order":6}],"host":"","image_url":"/api/forums/4157/cover"},{"id":4161,"title":"The 7th Chiplet Key Technologies Forum: Wafer-Level Systems","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"王郁杰","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-403","description":"Chiplet technology is pushing chip innovation from a single SoC toward system-level reconstruction. Compute, memory, I/O, and interconnect can be flexibly combined according to application requirements and integrated through advanced packaging processes to form larger-scale heterogeneous systems. As large model training and inference and high-performance computing continue to expand, traditional multi-chip board-level interconnection is gradually becoming limited in bandwidth, latency, power consumption, and density. Wafer-level computing has therefore become an important direction for chiplet technology to evolve toward higher integration. It incorporates chiplet interconnection, network-on-chip, power supply and cooling, yield redundancy, compilation mapping, and application workload co-design into a unified design, providing a new system path for computing power scaling in the post-Moore era.\nWhen the industry mentions wafer-level systems today, it mainly focuses on the packaging level: how to use advanced packaging technology to integrate multiple chiplets and other components together. Packaging companies and wafer foundries have done a lot of promotion in this regard. In the computer field, for system architecture designers, does this mean that organizing hardware together at the packaging layer is sufficient, and that migrating existing system design methods is enough? These are questions that need to be discussed.\nThe 7th Chiplet Key Technologies Forum takes \"wafer-level computing\" as its theme, focusing on wafer-level architecture, chiplet interconnection, power supply and cooling, compilation mapping, and system applications. It invites experts from academia and industry to exchange the latest progress, key challenges, and paths, and to promote the coordinated development of the wafer-level computing ecosystem.","is_demo":0,"cover_url":"/api/forums/4161/cover?size=thumb","cover_alt":"The 7th Chiplet Key Technologies Forum: Wafer-Level Systems","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4161","source_checked_at":"2026-10-30","agenda_json":"[[\"Thermal Management-Oriented Wafer-Scale Computing Optimization\",\"胡杨\"],[\"Mapping and Communication Optimization for Wafer-Scale Large Model Inference\",\"黄嘉逸\"],[\"Vertical Power Delivery for Wafer-Scale Systems: Challenges, Paths, and Breakthroughs\",\"尤祥安\"],[\"Large Model Wafer-Scale Systems: From Efficient Inference to Practical Deployment\",\"王梦迪\"]]","code":"CNCC2026-4161","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-403","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4161","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第七届芯粒关键技术论坛-晶圆级系统","original_topic":"新型计算与网络架构","original_description":"芯粒技术正在把芯片创新从单颗 SoC 推向系统级重构。计算、存储、I/O、互连可以按应用需求灵活组合，并通过先进封装集成工艺形成更大规模的异构系统。随着大模型训练推理和高性能计算持续扩大，传统多芯片板级互连在带宽、时延、功耗和密度上逐渐受限，晶圆级计算因此成为芯粒技术向更高集成度演进的重要方向。它将芯粒互连、片上网络、供电散热、良率冗余、编译映射和应用负载协同纳入统一设计，为后摩尔时代算力扩展提供新的系统路径。\n现在业界提到晶圆级系统主要聚焦于封装层面，如何使用先进封装技术将多个芯粒等部件集成在一起，封装厂、晶圆代工厂在这方面做了很多宣传。在计算机领域，对于系统架构设计人员来说，是不是意味着封装层讲硬件组织在一起就足够了，是不是将现有系统的设计方法迁移过来就够了，这都是需要探讨的问题。\n第七届芯粒关键技术论坛以“晶圆级计算”为主题，聚焦晶圆级架构、芯粒互联、供电散热、编译映射与系统应用，邀请产学研专家交流最新进展、关键挑战与路径，推动晶圆级计算生态协同发展。","people":[{"forum_id":4161,"role":"guest","name":"胡杨","affiliation":"清华大学","sort_order":0},{"forum_id":4161,"role":"chair","name":"王郁杰","affiliation":"中国科学院计算技术研究所","sort_order":0},{"forum_id":4161,"role":"guest","name":"黄嘉逸","affiliation":"香港科技大学（广州）","sort_order":1},{"forum_id":4161,"role":"guest","name":"尤祥安","affiliation":"中国科学院微电子研究所","sort_order":2},{"forum_id":4161,"role":"guest","name":"王梦迪","affiliation":"中国科学院计算技术研究所","sort_order":3}],"host":"","image_url":"/api/forums/4161/cover"},{"id":4164,"title":"4th Frontier Forum on Scientific Computing Platforms: AI4S Agent Exploration and Industry Empowerment","topic":"AI4Science","kind":"专题论坛","chair":"周号益","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-504","description":"Artificial intelligence is driving a paradigm reconstruction of scientific discovery and technological innovation. Recent advances such as Aletheia breaking through mathematical problems and AlphaEvolve discovering innovative algorithms confirm a new stage of \"autonomous discovery.\" Common scientific computing platforms support the coordinated evolution of entirely new theoretical paradigms, system technologies, and agent capabilities, leading the development of a new multi-disciplinary, large-scale scientific computing ecosystem. Currently, DeepMind, OpenAI, NVIDIA, xAI, and others are leveraging their AI software, hardware, and technical advantages in an effort to seize the initiative in this new field. Based on domestic AI foundational software and hardware, the CNAI4S AI scientific computing common platform continues to serve ubiquitous scientific domains. This forum invites top scholars in AI and scientific computing from home and abroad to give keynote reports, share technical practices represented by agents, and jointly discuss the new ecosystem for the large-scale growth of scientific computing.","is_demo":0,"cover_url":"/api/forums/4164/cover?size=thumb","cover_alt":"4th Frontier Forum on Scientific Computing Platforms: AI4S Agent Exploration and Industry Empowerment","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4164","source_checked_at":"2026-10-30","agenda_json":"[[\"Large-Small Model-Assisted Optimization of Complex Systems\",\"金耀初\"],[\"World Models and Cognitive Agents for Scientific Discovery\",\"吴文峻\"],[\"From Predicting Cells to Simulating Life: In Search of the \\\"Boltzmann Equation\\\" of Life\",\"彭玺\"],[\"AI for Scientific Simulation and Design\",\"朱霖潮\"],[\"The Development and Practice of AI4S in the Agentic Era\",\"张毅\"]]","code":"CNCC2026-4164","forum_type":"专题论坛","format":null,"theme_code":"AI4Science","theme_name":"AI4Science","venue":"西博城5号馆","room":"西博城5号馆-504","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4164","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第四届科学智算平台技术前沿：AI4S智能体探索与产业赋能","original_topic":"AI4Science","original_description":"人工智能正驱动科学发现与技术创新的范式重构，Aletheia突破数学难题、AlphaEvolve 发现创新算法等最新进展，印证着“自主发现”新阶段。科学智算共性平台支撑着全新理论范式、系统技术与智能体能力的协同演进，引领多学科规模化的科学智算新生态发展。目前，DeepMind、OpenAI、NVIDIA、xAI等正借助其人工智能软硬件与技术优势，力图在这一新领域占据发展先机。基于国产人工智能基础软硬件，CNAI4S人工智能科学计算共性平台持续服务泛在科学领域。本届论坛邀请来自国内外人工智能与科学计算领域顶尖学者进行主题报告，开展以智能体为代表的技术实践分享，共同研讨科学智算规模化生长新生态。","people":[{"forum_id":4164,"role":"chair","name":"周号益","affiliation":"北京航空航天大学","sort_order":0},{"forum_id":4164,"role":"guest","name":"金耀初","affiliation":"西湖大学","sort_order":0},{"forum_id":4164,"role":"guest","name":"吴文峻","affiliation":"北京航空航天大学","sort_order":1},{"forum_id":4164,"role":"guest","name":"彭玺","affiliation":"四川大学","sort_order":2},{"forum_id":4164,"role":"guest","name":"朱霖潮","affiliation":"浙江大学","sort_order":3},{"forum_id":4164,"role":"guest","name":"张毅","affiliation":"华为技术有限公司","sort_order":4}],"host":"","image_url":"/api/forums/4164/cover"},{"id":4174,"title":"The 3rd Intelligent Network and Network-Computing Integration Forum — In the Agent Era: How Can Networks and Computing Break Through and Reshape?","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"苏金树、陈星","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-502","description":"At present, global digital infrastructure is undergoing profound transformation and restructuring. Agents integrate environmental perception, memory reasoning, and tool invocation capabilities, driving the Internet to shift comprehensively from \"connecting people\" to \"connecting agents.\" It is expected that tens of billions of agents and new types of network applications will connect to networks, comprehensively reshaping the Internet's computing models, traffic patterns, communication paradigms, and trust systems. At the level of technological practice, the mutual constraint and bidirectional shaping of computing and networks are becoming increasingly prominent: on the one hand, in ten-thousand-card large model training scenarios, network communication accounts for as much as 30%-60% of the time, and network bandwidth, topology, and scheduling mechanisms have become core shortcomings constraining the release of computing power efficiency; on the other hand, with the rapid rise of ubiquitous computing and networking scenarios such as mixed deployment of large and small models, cloud-edge-device collaboration, satellite Internet, and the low-altitude economy, the scaled deployment of agents has demands on networks that go beyond traditional data transmission and computing power supply, requiring networks to simultaneously possess capabilities such as task understanding, service orchestration, collaborative control, and real-time response, and urgently requiring the deep integrated fusion of networks, computing, perception, and intelligence. The agent era is reshaping network forms and computing paradigms from the ground up—networks are no longer merely transmission pipelines, and computing is no longer limited to single-point servers; the two are deeply intertwined in every link of the agent lifecycle. Against this background, how networks and computing can break through and reshape has become a core proposition that the industry urgently needs to address.\n     In response to the above challenges and open questions, this forum invites several renowned experts and scholars to engage in in-depth deliberation on the opportunities and challenges faced by the integrated development of networks and computing in the agent era. This forum features three deliberative themes:\nDeliberation Point 1: The battle over architectural evolution—is it \"networks adapting to computing\" or \"computing reshaping networks\"?\nDeliberation Point 2: The change in interaction paradigms—should we pursue \"intent-driven ultimate network-computing efficiency\" or prioritize building a solid \"endogenous security value alignment defense line\"?\nDeliberation Point 3: The dilemma of deployment scenarios—for ubiquitous computing power and vertical scenarios, should it be a \"unified foundation\" or \"scenario customization\"?","is_demo":0,"cover_url":"/api/forums/4174/cover?size=thumb","cover_alt":"The 3rd Intelligent Network and Network-Computing Integration Forum — In the Agent Era: How Can Networks and Computing Break Through and Reshape?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4174","source_checked_at":"2026-10-30","agenda_json":"[[\"From the Innovation of the Open-Source Cybersecurity-Specific Programming System Yak to the Practice of the Industry-Grade Advanced Agent MemFit\",\"张小松\"],[\"A Brief Discussion on Cognitive Big Data Thinking in the Agent Era\",\"张清华\"],[\"Application Scenarios Driving the Development of Computing Power Network Technology\",\"窦勇\"],[\"Low-Altitude Intelligence and IoT\",\"李默\"],[\"Large-Small Model End-Cloud Collaborative Agent Systems\",\"郑臻哲\"]]","code":"CNCC2026-4174","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城5号馆","room":"西博城5号馆-502","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4174","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第三届智能网络与网算融合论坛 —— 智能体时代：网络与计算如何破局重塑？","original_topic":"新型计算与网络架构","original_description":"当前，全球数字基础设施正处于深刻变革与重构期。智能体融合环境感知、记忆推理与工具调用能力，推动互联网从“连接人”全面转向“连接智能体”，预计将有数以百亿计的智能体与新型网络应用接入网络，全方位重塑互联网的计算模式、流量形态、通信范式和信任体系。在技术实践层面，计算与网络相互制约、双向塑造的特征愈发凸显：一方面，在万卡级大模型训练场景中，网络通信耗时占比高达30%-60%，网络带宽、拓扑结构与调度机制已成为制约算力效能释放的核心短板；另一方面，随着大小模型混合部署、云边端协同、卫星互联网、低空经济等泛在计算与组网场景快速兴起，智能体的规模化部署对网络的诉求已超出传统数据传输与算力供给范畴，要求网络同时具备任务理解、服务编排、协同控制与实时响应等能力，亟需推动网络、计算、感知与智能的深度一体化融合。智能体时代正从底层重塑网络形态与计算范式——网络不再只是传输管道，计算也不再局限于单点服务器，二者在智能体生命周期的每一环节深度交织。在此背景下，网络与计算如何破局重塑，已成为行业亟待回应的核心命题。\n     针对上述挑战与开放问题，本论坛邀请多位知名专家学者，就智能体时代网络与计算融合发展面临的机遇与挑战展开深度思辨。本次论坛设置三大思辨主题：\n思辨点一：架构演进之争——是“网络适配计算”还是“计算重塑网络”？\n思辨点二：交互范式之变——是追求“意图驱动的极致网算效能”，还是优先筑牢“内生安全价值对齐防线”？\n思辨点三：落地场景之困——面向泛在算力与垂直场景，是“统一底座”还是“场景定制”？","people":[{"forum_id":4174,"role":"guest","name":"张小松","affiliation":"电子科技大学","sort_order":0},{"forum_id":4174,"role":"chair","name":"苏金树","affiliation":"军事科学院","sort_order":0},{"forum_id":4174,"role":"guest","name":"张清华","affiliation":"重庆邮电大学","sort_order":1},{"forum_id":4174,"role":"co_chair","name":"陈星","affiliation":"福州大学计算机与大数据学院","sort_order":1},{"forum_id":4174,"role":"guest","name":"窦勇","affiliation":"国防科学技术大学计算机学院","sort_order":2},{"forum_id":4174,"role":"guest","name":"李默","affiliation":"香港科技大学","sort_order":3},{"forum_id":4174,"role":"guest","name":"郑臻哲","affiliation":"上海交通大学","sort_order":4},{"forum_id":4174,"role":"guest","name":"王敬宇","affiliation":"北京邮电大学","sort_order":5},{"forum_id":4174,"role":"guest","name":"何倩","affiliation":"桂林电子科技大学","sort_order":6},{"forum_id":4174,"role":"guest","name":"赵宝康","affiliation":"国防科技大学","sort_order":7},{"forum_id":4174,"role":"guest","name":"陈哲毅","affiliation":"福州大学计算机与大数据学院","sort_order":8},{"forum_id":4174,"role":"guest","name":"迟国轩","affiliation":"北京邮电大学","sort_order":9},{"forum_id":4174,"role":"guest","name":"徐方鑫","affiliation":"云程半导体","sort_order":10}],"host":"","image_url":"/api/forums/4174/cover"},{"id":4176,"title":"From \"Generation\" to \"Production\": Will AI Programming Bring New \"Technical Debt\"?","topic":"AI Foundation Models","kind":"专题论坛","chair":"朱庆福、高翠芸","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-405","description":"AI programming is evolving from an auxiliary tool into autonomous agents with toolchain integration and specification-driven development, significantly improving code generation efficiency. However, the core of software engineering is not limited to code generation; it lies more in whether code can be continuously understood, verified, and maintained during the long-term evolution of systems. The continuous acceleration of code generation by AI programming may cause design drift, hidden coupling, and governance burdens masked by short-term efficiency or local usability to be introduced faster and continuously amplified, thereby forming new \"technical debt.\" To this end, this forum invites guests from academia and industry to deliberate on the following key challenges as AI programming moves from \"generation\" toward \"production\": 1. How should the boundaries of AI programming production capability be defined; 2. Within the capability boundaries, how can it adapt to continuously evolving system constraints and achieve long-term autonomy; 3. Beyond the capability boundaries, how can the contradiction between accelerated autonomous generation and limited manual review be coordinated to ensure trustworthy delivery.","is_demo":0,"cover_url":"/api/forums/4176/cover?size=thumb","cover_alt":"From \"Generation\" to \"Production\": Will AI Programming Bring New \"Technical Debt\"?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4176","source_checked_at":"2026-10-30","agenda_json":"[[\"On the Controllability of AI Programming\",\"金芝\"],[\"Opportunities and Challenges of Intelligent Software Development from a Complex Systems Perspective\",\"彭鑫\"],[\"From Programming Assistant to Digital Employee: The Next Stop for AI Coding\",\"谢吉宝\"],[\"New Infrastructure for Agent Training: Automated Code Execution Environment Construction System and Practice\",\"车万翔\"],[\"From Code Foundation Models to Code Agents\",\"杨健\"]]","code":"CNCC2026-4176","forum_type":"专题论坛","format":null,"theme_code":"AI基础模型","theme_name":"AI基础模型","venue":"西博城4号馆","room":"西博城4号馆-405","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4176","legacy_id":2,"lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"从“生成”到“生产”：AI 编程会带来新的“技术债”吗？","original_topic":"AI基础模型","original_description":"AI编程正在从辅助工具演化为具备工具链联动与规格驱动开发的自主代理，显著提升了代码生成效率。然而，软件工程的核心并不止于代码生成，更在于代码能否在系统长期演化中被持续理解、验证与维护。AI编程对代码生成的持续加速，可能使短期效率或局部可用性所掩盖的设计偏移、隐性耦合与治理负担被更快引入并持续放大，进而形成新的“技术债”。为此，本论坛邀请学界与业界嘉宾围绕 AI 编程从“生成”迈向“生产”的关键挑战开展如下思辨：1. AI 编程生产能力的边界如何界定；2. 在能力边界之内，如何适配持续演化的系统约束，实现长期自主；3. 在能力边界之外，如何协调自主生成加速与人工审查有限之间的矛盾，保障可信交付。","people":[{"forum_id":4176,"role":"chair","name":"朱庆福","affiliation":"哈尔滨工业大学","sort_order":0},{"forum_id":4176,"role":"guest","name":"金芝","affiliation":"北京大学","sort_order":0},{"forum_id":4176,"role":"co_chair","name":"高翠芸","affiliation":"哈尔滨工业大学（深圳）","sort_order":1},{"forum_id":4176,"role":"guest","name":"彭鑫","affiliation":"复旦大学","sort_order":1},{"forum_id":4176,"role":"guest","name":"谢吉宝","affiliation":"阿里巴巴","sort_order":2},{"forum_id":4176,"role":"guest","name":"车万翔","affiliation":"哈尔滨工业大学计算机科学与技术学院","sort_order":3},{"forum_id":4176,"role":"guest","name":"杨健","affiliation":"北京航空航天大学","sort_order":4},{"forum_id":4176,"role":"guest","name":"滕飞","affiliation":"西南交通大学","sort_order":5}],"host":"","image_url":"/api/forums/4176/cover"},{"id":4180,"title":"The 4th Edge Intelligence Technology Forum: Edge Intelligence Empowering \"AI+\"","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"何强","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-412","description":"With the in-depth advancement of the \"AI+\" initiative, artificial intelligence is accelerating its penetration into fields such as manufacturing, transportation, and urban governance. Facing needs such as real-time processing of massive data, limited computing power, privacy protection, and adaptation to complex scenarios, relying solely on cloud computing can no longer meet the needs of industrial intelligent development. Edge intelligence, by extending models, data, and computing power to the network edge and terminals, provides low-latency, highly reliable, and privacy-friendly technical support for \"AI+.\"\n\nThis forum has been successfully held at CNCC from 2023 to 2025. This year, it will discuss topics such as edge deployment of large models, device-edge-cloud collaboration, edge agents, and industrial applications, focusing on key issues such as limited edge computing power, efficient model adaptation, heterogeneous device collaboration, and industrial applications. It aims to bring together experts from industry, academia, and research to jointly explore the development path of edge intelligence empowering \"AI+\" and promote technological innovation and industrial application.","is_demo":0,"cover_url":"/api/forums/4180/cover?size=thumb","cover_alt":"The 4th Edge Intelligence Technology Forum: Edge Intelligence Empowering \"AI+\"","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4180","source_checked_at":"2026-10-30","agenda_json":"[[\"Device-Side Sensing and Edge Intelligence Empowered Mobile Ubiquitous Healthcare\",\"薛广涛\"],[\"New Developments in Industrial Cloud-Edge Intelligence: Industrial Large Models + Industrial Agents\",\"任磊\"],[\"Model Internet\",\"李哲涛\"],[\"Key Technologies of Intelligent Infrared Satellites\",\"孔令和\"],[\"Key Technologies of Green Computing Power and Large Model Scheduling\",\"邱铁\"]]","code":"CNCC2026-4180","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-412","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4180","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"第四届边缘智能技术论坛：边缘智能赋能“人工智能+”","original_topic":"新型计算与网络架构","original_description":"随着“人工智能+”行动深入推进，人工智能正加速向制造、交通和和城市治理等领域渗透。面对海量数据实时处理、算力受限、隐私保护及复杂场景适配等需求，单纯依赖云端计算已难以满足产业智能化发展需要。边缘智能通过推动模型、数据与算力向网络边缘和终端延伸，为“人工智能+”提供低时延、高可靠、隐私友好的技术支撑。\n\n本论坛已于2023-2025年在CNCC成功举办，今年将围绕大模型边缘部署、端边云协同、边缘智能体及产业应用等议题展开讨论，重点关注边缘算力受限、模型高效适配、异构设备协同和产业应用等关键问题。旨在汇聚产学研专家，共同探讨边缘智能赋能“人工智能+”的发展路径，推动技术创新与产业应用。","people":[{"forum_id":4180,"role":"chair","name":"何强","affiliation":"华中科技大学","sort_order":0},{"forum_id":4180,"role":"guest","name":"薛广涛","affiliation":"上海交通大学","sort_order":0},{"forum_id":4180,"role":"guest","name":"任磊","affiliation":"北京航空航天大学","sort_order":1},{"forum_id":4180,"role":"guest","name":"李哲涛","affiliation":"暨南大学","sort_order":2},{"forum_id":4180,"role":"guest","name":"孔令和","affiliation":"上海交通大学","sort_order":3},{"forum_id":4180,"role":"guest","name":"邱铁","affiliation":"天津大学","sort_order":4}],"host":"","image_url":"/api/forums/4180/cover"},{"id":4182,"title":"Building a New-Generation Operating System AgentOS: From Technological Breakthroughs to Industrial Deployment","topic":"AI Agents","kind":"专题论坛","chair":"朱少民","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-501","description":"What will be discussed? The forum will center on the concept, overall architecture, and key technologies of AgentOS, with emphasis on agent lifecycle management, task scheduling, multi-agent collaboration, memory and knowledge retrieval, tool invocation, permission control, security sandboxes, human-machine collaboration, and system observability, and will analyze the role of open protocols such as MCP and A2A.\nWhat questions will be raised? How to effectively manage new types of resources such as Tokens, context, and API quotas; how to improve the reliability, interpretability, and fault tolerance of multi-agent collaboration; how to prevent unauthorized operations, data leakage, and prompt injection; and how to establish unified standards and evaluation systems.\nWhy recommend this topic? AgentOS is driving AI from a \"single-point question-answering tool\" toward a collaborative and governable \"digital employee organization,\" and is key infrastructure for the scaled deployment of large models. This topic combines frontier relevance, technical challenges, and industrial value.","is_demo":0,"cover_url":"/api/forums/4182/cover?size=thumb","cover_alt":"Building a New-Generation Operating System AgentOS: From Technological Breakthroughs to Industrial Deployment","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4182","source_checked_at":"2026-10-30","agenda_json":"[[\"Lingji OS: Breaking Agent Silos and Building an Open, Interconnected Open-Source Agent Operating System\",\"谢涛\"],[\"Architectural Reflections and Ecosystem Practices of the Nuwa Agent OS\",\"赵宇\"],[\"PilotDeck — A Productivity-Oriented Agent Operating System Attempt\",\"闫宇坤\"],[\"Exploration and Practice of DevStar AI's AgentOS in Software Engineering\",\"孟宁\"],[\"AgentOS: A Native Operating System Foundation Supporting Continuous Evolution of Long-Horizon Agent Tasks\",\"喜超\"],[\"Panel Host: AgentOS: An Inevitable Choice or an Optional Solution for Enterprise Intelligence?\",\"朱少民\"]]","code":"CNCC2026-4182","forum_type":"专题论坛","format":null,"theme_code":"Agent智能体","theme_name":"Agent智能体","venue":"西博城5号馆","room":"西博城5号馆-501","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4182","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"构建新一代操作系统AgentOS：从技术突破到产业落地","original_topic":"Agent智能体","original_description":"讨论哪些内容？围绕AgentOS的概念、总体架构和关键技术展开，重点讨论智能体生命周期管理、任务调度、多智能体协作、记忆与知识检索、工具调用、权限控制、安全沙箱、人机协同及系统可观测性，并分析MCP、A2A 等开放协议的作用。\n提出哪些问题？探讨如何有效管理Token、上下文和 API 配额等新型资源；如何提升多智能体协作的可靠性、可解释性与容错能力；如何防范越权操作、数据泄露和提示注入；以及如何建立统一标准和评估体系。\n为何推荐此选题？AgentOS正推动AI从“单点问答工具”走向可协作、可治理的“数字员工组织”，是大模型规模化落地的关键基础设施。该选题兼具前沿性、技术挑战和产业价值。","people":[{"forum_id":4182,"role":"chair","name":"朱少民","affiliation":"同济大学软件学院","sort_order":0},{"forum_id":4182,"role":"guest","name":"谢涛","affiliation":"北京大学","sort_order":0},{"forum_id":4182,"role":"guest","name":"赵宇","affiliation":"西南财经大学","sort_order":1},{"forum_id":4182,"role":"guest","name":"闫宇坤","affiliation":"启元实验室","sort_order":2},{"forum_id":4182,"role":"guest","name":"孟宁","affiliation":"梦宁软件","sort_order":3},{"forum_id":4182,"role":"guest","name":"喜超","affiliation":"华为云技术公司","sort_order":4}],"host":"","image_url":"/api/forums/4182/cover"},{"id":4183,"title":"Seventy Years of Legacy, a New Era of Intelligence: Rereading China's Computer History in the AI Era","topic":"Unclassified","kind":"专题论坛","chair":"赵轲","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城2号馆-208","description":"As a signature event of the CCF Computer History Working Committee, this forum will invite experts in computer history, technology, and industry to look back on the journey at the 70-year historical milestone, from early computer development to the rise of contemporary artificial intelligence, to discuss the historical trajectory and spiritual inheritance of China's computer endeavor, and to look ahead to the new missions and new possibilities of computer history research in the intelligent era. Standing at this special point in time, we must not only remember those pioneers who blazed the trail through hardships, but also, today when artificial intelligence is profoundly changing computing paradigms and the industrial landscape, it is all the more necessary to re-examine the contemporary significance of computer history. History has never been so urgently in need of interpretation—because only by understanding \"where we came from\" can we more clearly judge \"where we are going.\"","is_demo":0,"cover_url":"/api/forums/4183/cover?size=thumb","cover_alt":"Seventy Years of Legacy, a New Era of Intelligence: Rereading China's Computer History in the AI Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4183","source_checked_at":"2026-10-30","agenda_json":"[[\"The 441-B Transistor Computer of the Harbin Military Engineering Institute\",\"徐祖哲\"],[\"Hard-to-Compute Contributions: An Overview of CCF's Activities in Recording Chinese Computer History\",\"司宏伟\"],[\"Another Possibility for Computing History: The Technical Lineage of the MONIAC Hydraulic Analog Computer\",\"刘年凯\"],[\"Seventy Years of Artificial Intelligence\",\"李晶晶\"],[\"Preserving Memory Through Repetition: A History of Computer Systems from Data Compression to Deduplication\",\"李经纬\"]]","code":"CNCC2026-4183","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城2号馆","room":"西博城2号馆-208","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4183","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"薪火七十载，智启新纪元：人工智能时代重读中国计算机史","original_topic":"未分类","original_description":"本论坛作为CCF计算机历史工委的品牌活动，将邀请计算机历史、技术、产业领域专家，在70年历史节点上回望来路，从早期计算机研制到当代人工智能崛起，探讨中国计算机事业发展的历史脉络与精神传承，并展望智能时代计算机历史研究的新使命与新可能。站在这个特殊的时间节点上，我们不仅要铭记那些筚路蓝缕的开拓者，更有必要在人工智能深刻改变计算范式与产业格局的今天，重新审视计算机历史的当代意义。历史从未如此迫切地需要被解读——因为理解“我们从哪里来”，才能更清醒地判断“我们向何处去”。","people":[{"forum_id":4183,"role":"chair","name":"赵轲","affiliation":"电子科技大学","sort_order":0},{"forum_id":4183,"role":"guest","name":"徐祖哲","affiliation":"数据通信科学技术研究所","sort_order":0},{"forum_id":4183,"role":"guest","name":"司宏伟","affiliation":"中国科学院自然科学史研究所","sort_order":1},{"forum_id":4183,"role":"guest","name":"刘年凯","affiliation":"清华大学","sort_order":2},{"forum_id":4183,"role":"guest","name":"李晶晶","affiliation":"电子科技大学","sort_order":3},{"forum_id":4183,"role":"guest","name":"李经纬","affiliation":"电子科技大学","sort_order":4}],"host":"","image_url":"/api/forums/4183/cover"},{"id":4187,"title":"Human-Machine-Environment Integration and Coevolution in the Agent Era","topic":"Unclassified","kind":"专题论坛","chair":"周晟、刘佐珠","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-508","description":"Agents are changing from \"tools that can converse\" into entities that continuously act in real environments: they call software interfaces, drive robot bodies, embed themselves in organizational processes, and continuously self-evolve after deployment. The traditional \"human-computer interaction\" framework is therefore facing failure. This forum brings humans, machines, and environments into the same framework and conducts speculation around three topics:\n\n1. The struggle for adaptation: Do agents adapt to the world, or is the world reconstructed for agents?\n2. The evaluation dilemma: Can continuously evolving agents still be measured accurately?\n3. The path to co-progress: In human-machine coevolution, how can people become stronger the more they use it?\n\nThe forum covers three types of environments — digital, physical, and social organizational — and strives to form actionable judgments and consensus recommendations.","is_demo":0,"cover_url":"/api/forums/4187/cover?size=thumb","cover_alt":"Human-Machine-Environment Integration and Coevolution in the Agent Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4187","source_checked_at":"2026-10-30","agenda_json":"[[\"World Modeling, Generation, and Interaction\",\"齐晓娟\"],[\"Endless Sound: A Preliminary Study of Acoustic Resonance-Based Fine-Grained Liquid Detection\",\"张永敏\"],[\"Perception, Reasoning, and Interaction in Audio Embodied Intelligence\",\"刘李\"],[\"Agent Research and Applications for Complex Scenarios\",\"高莘\"]]","code":"CNCC2026-4187","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城5号馆","room":"西博城5号馆-508","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4187","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智能体时代的人-机-环融合与协同进化","original_topic":"未分类","original_description":"智能体正从\"能对话的工具\"变为在真实环境中持续行动的主体：它调用软件接口、驱动机器人本体、嵌入组织流程，并在部署后不断自我进化。传统\"人机交互\"框架因此面临失效。本论坛将人、机、环纳入同一框架，围绕三个议题展开思辨：\n\n1. 适应之争：智能体适应世界，还是世界为智能体重构？\n2. 评测之困：持续进化的智能体，还测得准吗？\n3. 共进之道：人机协同进化，如何让人越用越强？\n\n论坛覆盖数字、物理与社会组织三类环境，力求形成可落地的判断与共识建议。","people":[{"forum_id":4187,"role":"chair","name":"周晟","affiliation":"浙江大学","sort_order":0},{"forum_id":4187,"role":"guest","name":"齐晓娟","affiliation":"香港大学","sort_order":0},{"forum_id":4187,"role":"guest","name":"张永敏","affiliation":"中南大学","sort_order":1},{"forum_id":4187,"role":"co_chair","name":"刘佐珠","affiliation":"浙江大学","sort_order":1},{"forum_id":4187,"role":"guest","name":"刘李","affiliation":"香港科技大学（广州）","sort_order":2},{"forum_id":4187,"role":"guest","name":"高莘","affiliation":"电子科技大学","sort_order":3}],"host":"","image_url":"/api/forums/4187/cover"},{"id":4192,"title":"Transforming Computer Education in the AI Era","topic":"General","kind":"专题论坛","chair":"张莉","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城2号馆-204","description":"Artificial intelligence technologies represented by large models and agents are reshaping every industry at an unprecedented pace, and computer education stands at a critical crossroads in its leap from \"digitalization\" to \"intelligentization.\" This forum aims to bring together top forces from academia, industry, and research, and adopts a structure of \"frontier reports + in-depth deliberation (Panel)\" to directly confront the dual effects of \"empowerment\" and \"disempowerment\" that AI technology brings to computer education, to explore how to reconstruct the talent cultivation system, innovate teaching models, and reshape the open-source and engineering ecosystem, and to provide theoretical support and practical paradigms for the high-quality development of China's computer education during the \"15th Five-Year Plan\" period.","is_demo":0,"cover_url":"/api/forums/4192/cover?size=thumb","cover_alt":"Transforming Computer Education in the AI Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4192","source_checked_at":"2026-10-30","agenda_json":"[[\"Agent-Driven Smart Metaverse Education and Competency Development\",\"徐晓飞\"],[\"Opportunities and Challenges in Software Talent Cultivation in the Intelligent Era\",\"李宣东\"],[\"Challenges and Transformations in Software Talent Cultivation in the AI Era\",\"毛新军\"],[\"Competency Restructuring for Computing Talent in the Era of Large Models: Industry-Perspective Assessment Observations and Tiered Cultivation Practices\",\"林华鼎\"]]","code":"CNCC2026-4192","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城2号馆","room":"西博城2号馆-204","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4192","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI时代的计算机教育变革","original_topic":"综合","original_description":"以大模型和智能体（Agent）为代表的人工智能技术正以前所未有的速度重塑千行百业，计算机教育正处于从“数字化”向“数智化”跃迁的关键十字路口，本论坛旨在汇聚产学研顶尖力量，，采用“前沿报告+深度思辨（Panel）”的结构，直面AI技术对计算机教育带来的“赋能”与“负能”双重效应，探讨如何重构人才培养体系、创新教学模式以及重塑开源与工程生态，为“十五五”期间我国计算机教育的高质量发展提供理论支撑与实践范式。","people":[{"forum_id":4192,"role":"guest","name":"徐晓飞","affiliation":"哈尔滨工业大学","sort_order":0},{"forum_id":4192,"role":"chair","name":"张莉","affiliation":"北京航空航天大学","sort_order":0},{"forum_id":4192,"role":"guest","name":"李宣东","affiliation":"南京大学计算机科学与技术系","sort_order":1},{"forum_id":4192,"role":"guest","name":"毛新军","affiliation":"国防科技大学","sort_order":2},{"forum_id":4192,"role":"guest","name":"林华鼎","affiliation":"华为技术有限公司","sort_order":3}],"host":"","image_url":"/api/forums/4192/cover"},{"id":4193,"title":"Smart Party Building: Agent-Driven Organizational Transformation, Governance Innovation, and Value Alignment","topic":"AI for Industry","kind":"专题论坛","chair":"余霄","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-512","description":"On June 15, the National Party Building Work Symposium was held, making work arrangements for studying and implementing Xi Jinping's thought on Party building. The General Secretary has always emphasized that the application of information technology is an inevitable requirement for the reform and innovation of Party building work, and that it is necessary to proactively adapt to the new situation of the information age and the new changes in Party member teams, actively use emerging technologies, innovate the content and methods of Party organization activities, and advance smart Party building. The core topic of this forum is the technical architecture and trustworthy governance mechanisms of agent-driven Party building work, including knowledge base construction, content provenance, hallucination suppression, and value alignment. The reason for recommendation lies in its distinctive interdisciplinary character, connecting computer science and Party building, integrating the rationality of technical tools with the rationality of organizational values. It not only responds to the strategic deployment of the National Party Building Work Symposium but also echoes the theme of CNCC on gathering data to empower industry and computing intelligence for the future, and has important academic value and practical significance for promoting smart Party building from concept to practice and from pilot to standardization.","is_demo":0,"cover_url":"/api/forums/4193/cover?size=thumb","cover_alt":"Smart Party Building: Agent-Driven Organizational Transformation, Governance Innovation, and Value Alignment","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4193","source_checked_at":"2026-10-30","agenda_json":"[[\"Party Affairs Agents: Paradigm Construction, Value Alignment, and Grassroots Practice Paths from the Perspective of Humanities-Engineering Interdisciplinarity\",\"余霄\"],[\"Big Data-Driven Party Building-Led Grassroots Governance: Efficiency Improvement, Data Compliance, and Practice Paths\",\"周涛\"]]","code":"CNCC2026-4193","forum_type":"专题论坛","format":null,"theme_code":"AI+行业","theme_name":"AI+行业","venue":"西博城5号馆","room":"西博城5号馆-512","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":2,"forum_id":"CNCC2026-4193","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智慧党建：智能体驱动的组织变革、治理创新与价值对齐","original_topic":"AI+行业","original_description":"6月15日全国党建工作座谈会召开，对学习贯彻习近平党建思想作出工作部署。总书记历来强调信息技术应用是党建工作改革创新的必然要求，要主动适应信息时代新形势和党员队伍新变化，积极运用新兴技术，创新党组织活动内容方式，推进智慧党建。本论坛核心议题是智能体驱动党建工作的技术架构与可信治理机制，包括知识库构建、内容溯源、幻觉抑制、价值对齐等。推荐理由在于其鲜明的交叉学科特质，连接计算机科学与党的建设，融合技术工具理性与组织价值理性，既响应全国党建工作座谈会的战略部署，又呼应CNCC数聚产业、智算未来大会主题，对推动智慧党建从概念走向实践、从试点走向规范具有重要的学术价值与现实意义。","people":[{"forum_id":4193,"role":"chair","name":"余霄","affiliation":"四川省委党校 四川行政学院","sort_order":0},{"forum_id":4193,"role":"guest","name":"张婧","affiliation":"四川明炬律师事务所","sort_order":0},{"forum_id":4193,"role":"guest","name":"袁宏伟","affiliation":"中共四川省委党校","sort_order":1},{"forum_id":4193,"role":"guest","name":"苏自然","affiliation":"国家大数据安全靶场应用场景创新基地","sort_order":2},{"forum_id":4193,"role":"guest","name":"刘古权","affiliation":"致远协同研究院","sort_order":3},{"forum_id":4193,"role":"guest","name":"韩丽敏","affiliation":"中国石油“互联网+国企党建”研究院","sort_order":4},{"forum_id":4193,"role":"guest","name":"裴泽庆","affiliation":"中共四川省委党校","sort_order":5},{"forum_id":4193,"role":"guest","name":"郭从伦","affiliation":"中共四川省委党校","sort_order":6},{"forum_id":4193,"role":"guest","name":"周涛","affiliation":"电子科技大学","sort_order":7}],"host":"","image_url":"/api/forums/4193/cover"},{"id":4196,"title":"APAC Flagship Forum — The 10th Asia-Pacific Forum on AI for Aging and Disability Support","topic":"Unclassified","kind":"专题论坛","chair":"陈益强","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"天府国际会议中心-110卧龙","description":"The Asia-Pacific Association for Computer Science (APAC) Flagship Forum — The 10th Asia-Pacific Forum on AI for Aging and Disability Support, as a featured forum of the CNCC conference, brings together renowned experts and scholars from well-known universities and research institutions across Asia-Pacific countries and regions. The forum conducts in-depth discussion around the development status and future trends of information technology empowering aging and disability support and the healthcare field, focuses on cutting-edge technologies such as AI large models, affective computing, and federated learning, and explores relying on collaborative intelligence to create new-quality service capabilities and build an inclusive, intelligent smart assistance service system for older adults and people with disabilities.","is_demo":0,"cover_url":"/api/forums/4196/cover?size=thumb","cover_alt":"APAC Flagship Forum — The 10th Asia-Pacific Forum on AI for Aging and Disability Support","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4196","source_checked_at":"2026-10-30","agenda_json":"[[\"Opening Remarks\",\"杜子德\"],[\"AI-Empowered Healthy Aging and Longevity\",\"苗春燕\"],[\"Artificial Intelligence and Mental Health\",\"张震\"],[\"Beyond Accessibility: Leveraging AI to Expand Digital and Physical Environment Accessibility for People with Visual Impairments\",\"Uran Oh\"],[\"From Connectivity to Collaboration: A New Paradigm of Inclusive Services for the Model Internet\",\"凤阳\"],[\"From \\\"Governance\\\" to \\\"Intelligent Governance\\\": Infectious Disease Prevention and Control Empowered by Data Science and Artificial Intelligence\",\"刘泱\"],[\"Ten-Year Summary of the Asia-Pacific Forum on Aging and Disability Technologies\",\"陈益强\"]]","code":"CNCC2026-4196","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"天府国际会议中心","room":"天府国际会议中心-110卧龙","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":7,"forum_id":"CNCC2026-4196","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"亚太计算机学会联合会（APAC）旗舰论坛——第十届亚太AI适老助残技术论坛","original_topic":"未分类","original_description":"亚太计算机学会联合会（APAC）旗舰论坛 —— 第十届亚太 AI 适老助残技术论坛，作为 CNCC 大会特色论坛，汇聚亚太各国及地区知名高校、科研机构领域专家学者。论坛围绕信息技术赋能适老助残、健康医疗领域的发展现状与未来趋势展开深度研讨，聚焦 AI 大模型、情感计算、联邦学习等前沿技术，探索依托协同智能打造新质服务能力，面向老年人与残障群体构建普惠、智能的智慧辅助服务体系。","people":[{"forum_id":4196,"role":"chair","name":"陈益强","affiliation":"中国科学院计算技术研究所","sort_order":0},{"forum_id":4196,"role":"guest","name":"杨强","affiliation":"香港理工大学AI高等研究院","sort_order":0},{"forum_id":4196,"role":"guest","name":"杜子德","affiliation":"香港中文大学（深圳）","sort_order":1},{"forum_id":4196,"role":"guest","name":"苗春燕","affiliation":"新加坡南洋理工大学","sort_order":2},{"forum_id":4196,"role":"guest","name":"张震","affiliation":"悉尼中央精神科诊所","sort_order":3},{"forum_id":4196,"role":"guest","name":"Uran Oh","affiliation":"Ewha Womans University","sort_order":4},{"forum_id":4196,"role":"guest","name":"凤阳","affiliation":"中国电信集团教育行业事业部","sort_order":5},{"forum_id":4196,"role":"guest","name":"刘泱","affiliation":"香港浸会大学","sort_order":6}],"host":"","image_url":"/api/forums/4196/cover"},{"id":4200,"title":"One-Person Companies in the AI Era: Where Do Superpowers Come From and Where Are They Heading?","topic":"Unclassified","kind":"专题论坛","chair":"杨晓晖、彭锦佳","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-408","description":"Generative AI is reshaping the productivity paradigm and giving rise to AI-empowered one-person companies (OPCs), a new form of innovation and entrepreneurship. Unlike traditional small and micro businesses, OPCs take the individual as the hub and AI as the tool to achieve efficient one-person commercialization. With the implementation of relevant policies and industry standards, the OPC industry is gradually becoming standardized, but many issues remain to be clarified: how to define the core differences between OPCs and traditional entrepreneurial models and their long-term evolutionary forms; as AI lowers the execution threshold, do practitioners' capability requirements rise or fall, and what core capabilities should practitioners possess; and how to build a stable, compliant, replicable, and sustainable OPC commercial operating model. Based on a cross-sector perspective of industry, academia, and research, this forum integrates theoretical research and industrial practice for in-depth discussion, aiming to clarify the core nature and industry boundaries of the OPC business form, explore a development path that balances innovation vitality and compliance governance, and distill a development paradigm that can be promoted across industries, providing reference for policy optimization, entrepreneurial incubation, and industrial ecosystem building.","is_demo":0,"cover_url":"/api/forums/4200/cover?size=thumb","cover_alt":"One-Person Companies in the AI Era: Where Do Superpowers Come From and Where Are They Heading?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4200","source_checked_at":"2026-10-30","agenda_json":"[[\"The Rise of the Super Individual: Value Creation and Reconstruction of Production Relations in the AI Era\",\"张莹\"],[\"AI Thinking in the OPC Era\",\"王涛\"],[\"Business Launch on Day One: OPC-Oriented AI Application Construction Methods\",\"范唯媛\"],[\"Letting Professional Capabilities Break Through Human Boundaries: OPC Practice and Reflections of Fasuo AI\",\"程子洋\"]]","code":"CNCC2026-4200","forum_type":"专题论坛","format":null,"theme_code":null,"theme_name":null,"venue":"西博城4号馆","room":"西博城4号馆-408","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[\"theme\"]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4200","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI时代一人公司：超能力从何而来，路向何方？","original_topic":"未分类","original_description":"生成式AI重构生产力范式，催生AI赋能型一人公司（OPC）这一全新创新创业形态。区别于传统小微业态，OPC以个体为中枢、AI为工具，实现单人高效商业化落地。伴随相关政策与行业标准落地，OPC产业逐步规范化发展，但仍存在诸多议题有待明确：如何界定OPC与传统创业模式的核心差异及其长期演化形态；AI降低执行门槛的同时，从业者能力要求是上升还是下降，从业者应当具备哪些核心能力；如何构建稳定合规、可复制、可持续的OPC商业化运营模式。本论坛立足产学研跨界视角，融合理论研究与产业实践经验开展深度研讨，旨在厘清OPC业态内核与行业边界，探索平衡创新活力与合规治理的发展路径，提炼可跨行业推广的发展范式，为政策优化、创业孵化、产业生态建设提供参考。","people":[{"forum_id":4200,"role":"guest","name":"张莹","affiliation":"华北电力大学","sort_order":0},{"forum_id":4200,"role":"chair","name":"杨晓晖","affiliation":"河北大学","sort_order":0},{"forum_id":4200,"role":"co_chair","name":"彭锦佳","affiliation":"河北大学","sort_order":1},{"forum_id":4200,"role":"guest","name":"王涛","affiliation":"星科智汇（苏州）科技有限公司","sort_order":1},{"forum_id":4200,"role":"guest","name":"范唯媛","affiliation":"阿里云计算有限公司","sort_order":2},{"forum_id":4200,"role":"guest","name":"程子洋","affiliation":"法索 AI","sort_order":3},{"forum_id":4200,"role":"guest","name":"黄子群","affiliation":"OPC 圈（北京数据胶囊科技有限公司）","sort_order":4},{"forum_id":4200,"role":"guest","name":"黄凯","affiliation":"北京百度网讯科技有限公司","sort_order":5}],"host":"","image_url":"/api/forums/4200/cover"},{"id":4203,"title":"AI Diving Deep into Bioinformatics: Decoding Life or Creating More Precise Black Boxes?","topic":"AI4Science","kind":"专题论坛","chair":"王爽、辛鑫","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-515","description":"AI technologies represented by AlphaFold, ESM, and single-cell multi-omics large models are deeply intervening in bioinformatics. AI has expanded from a data analysis tool to an important force in hypothesis generation, mechanism inference, and experimental design, making continuous breakthroughs in tasks such as protein structure prediction, gene regulation inference, and drug target discovery. At the same time, the epistemological risks behind this prosperity have become more prominent: do the patterns recognized by models correspond to real life mechanisms, or are they statistical correlations in high-dimensional data? As researchers increasingly rely on AI outputs for scientific judgment, are we accelerating our understanding of the complexity of life, or are we handing over the right of interpretation to more precise yet more unknowable black boxes?\nThis forum focuses on the mechanism explanation, trustworthy validation, and technical route issues raised by AI diving deep into bioinformatics, and discusses how to establish a sustainable scientific paradigm among predictive capability, mechanistic understanding, and experimental validation.","is_demo":0,"cover_url":"/api/forums/4203/cover?size=thumb","cover_alt":"AI Diving Deep into Bioinformatics: Decoding Life or Creating More Precise Black Boxes?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4203","source_checked_at":"2026-10-30","agenda_json":"[[\"From AI4S to AI Scientist: How AI is revolutionizing drug discovery?\",\"洪鑫\"],[\"Genomic Language Models with Dual-Chain Collaborative Modeling\",\"刘元盛\"],[\"Natural Language Understanding-Based Protein Mining and Design System\",\"原发杰\"],[\"SPDEBench: A Benchmark for Learning Regular and Singular Stochastic PDEs\",\"孟琪\"]]","code":"CNCC2026-4203","forum_type":"专题论坛","format":null,"theme_code":"AI4Science","theme_name":"AI4Science","venue":"西博城5号馆","room":"西博城5号馆-515","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4203","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"AI深潜生物信息学，是在解码生命还是制造更精密的黑盒？","original_topic":"AI4Science","original_description":"以AlphaFold、ESM和单细胞多组学大模型为代表的AI技术，正深度介入生物信息学。AI已从数据分析工具扩展为假说生成、机制推断与实验设计的重要力量，在蛋白质结构预测、基因调控推断、药物靶点发现等任务中持续突破。与此同时，繁荣背后的认识论风险更加凸显：模型识别的模式究竟对应真实生命机制，还是高维数据中的统计相关？当研究者越来越依赖AI输出进行科学判断，我们是在加速理解生命复杂性，还是把解释权交给更精密却更不可知的黑盒？\n本论坛聚焦AI深潜生物信息学引发的机理解释、可信验证与技术路线问题，讨论如何在预测能力、机制理解和实验验证之间建立可持续的科学范式。","people":[{"forum_id":4203,"role":"chair","name":"王爽","affiliation":"中国石油大学（华东）","sort_order":0},{"forum_id":4203,"role":"guest","name":"洪鑫","affiliation":"清华大学智能产业研究院（AIR）","sort_order":0},{"forum_id":4203,"role":"co_chair","name":"辛鑫","affiliation":"山东大学","sort_order":1},{"forum_id":4203,"role":"guest","name":"刘元盛","affiliation":"湖南大学","sort_order":1},{"forum_id":4203,"role":"guest","name":"原发杰","affiliation":"西湖大学","sort_order":2},{"forum_id":4203,"role":"guest","name":"孟琪","affiliation":"中国科学院数学与系统科学研究院","sort_order":3}],"host":"","image_url":"/api/forums/4203/cover"},{"id":4208,"title":"Intelligent Storage Integration: Large Model Storage Technology Forum","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"吴晨涛","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-413","description":"This forum focuses on the evolution and innovation of storage technology in the era of large models, and conducts in-depth exchanges on key topics such as vector databases, distributed storage, hot and cold data tiering, storage-compute collaboration, data lakehouse, and AI infrastructure optimization. As the parameter scale, training corpora, and inference access volumes of large models continue to grow, storage systems are shifting from traditional capacity support to a high-performance, low-latency, highly reliable, and intelligent data foundation. This forum will invite experts, scholars, and industry representatives from universities, research institutions, cloud computing enterprises, storage vendors, and large model application teams to share cutting-edge research results, engineering practice experience, and judgments on future trends, and to discuss how to improve large model training and inference efficiency, reduce data movement and system costs through \"intelligent storage integration,\" and promote the development of AI infrastructure toward greater efficiency, sustainability, and trustworthiness. The forum aims to build a collaborative exchange platform for industry, academia, research, and application, and to promote breakthroughs in key technologies and ecosystem co-construction for large model storage.","is_demo":0,"cover_url":"/api/forums/4208/cover?size=thumb","cover_alt":"Intelligent Storage Integration: Large Model Storage Technology Forum","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4208","source_checked_at":"2026-10-30","agenda_json":"[[\"High-Performance Storage Technology for Supercomputing-Intelligence Integration\",\"肖侬\"],[\"Research on Conditional Filtering Vector Retrieval Technology Adapted to Business Workloads\",\"张广艳\"],[\"Efficient Data Storage and Resource Management for Intelligent Large Model Inference\",\"李永坤\"],[\"Research on High-Efficiency Storage Technology for Intelligence-Storage Integration Scenarios\",\"夏文\"],[\"Research on Efficient Heterogeneous Memory Data Deduplication and Management Technology\",\"毛波\"],[\"TensorDS: A Tensor Data System\",\"曾令仿\"]]","code":"CNCC2026-4208","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-413","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4208","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"智存融合—大模型存储技术论坛","original_topic":"新型计算与网络架构","original_description":"本论坛聚焦大模型时代存储技术的演进与创新，围绕向量数据库、分布式存储、冷热数据分层、存算协同、数据湖仓与AI基础设施优化等关键议题展开深入交流。随着大模型参数规模、训练语料和推理访问量持续增长，存储系统正从传统容量支撑转向高性能、低延迟、高可靠与智能化的数据底座。本论坛将邀请来自高校、科研机构、云计算企业、存储厂商及大模型应用团队的专家学者和产业代表，分享前沿研究成果、工程实践经验与未来趋势判断，探讨如何通过“智存融合”提升大模型训练与推理效率，降低数据流动和系统成本，推动AI基础设施向更高效、更绿色、更可信方向发展。论坛旨在搭建产学研用协同交流平台，促进大模型存储关键技术突破与生态共建。","people":[{"forum_id":4208,"role":"chair","name":"吴晨涛","affiliation":"上海交通大学","sort_order":0},{"forum_id":4208,"role":"guest","name":"谷云飞","affiliation":"上海交通大学","sort_order":0},{"forum_id":4208,"role":"guest","name":"肖侬","affiliation":"中山大学计算机学院","sort_order":1},{"forum_id":4208,"role":"guest","name":"张广艳","affiliation":"清华大学","sort_order":2},{"forum_id":4208,"role":"guest","name":"李永坤","affiliation":"中国科学技术大学","sort_order":3},{"forum_id":4208,"role":"guest","name":"夏文","affiliation":"哈尔滨工业大学（深圳）","sort_order":4},{"forum_id":4208,"role":"guest","name":"毛波","affiliation":"厦门大学","sort_order":5},{"forum_id":4208,"role":"guest","name":"曾令仿","affiliation":"之江实验室","sort_order":6}],"host":"","image_url":"/api/forums/4208/cover"},{"id":4215,"title":"Steering Agent Systems: Security Dilemmas and Trust Reconstruction of Agent OS","topic":"AI Safety and Governance","kind":"专题论坛","chair":"张殷乾","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-505","description":"As artificial intelligence evolves from single-turn interaction to an agent form with autonomous decision-making, complex planning, and cross-environment execution capabilities, the agent operating system (Agent OS), as a new-generation computing foundation, is fundamentally reconstructing the basic logic of software operation. However, the inherent autonomy and high dynamism of agents have dissolved the security boundaries and isolation mechanisms established by traditional operating systems relying on deterministic code and static privileges, thereby giving rise to multi-dimensional system-level security dilemmas, including loss of control over nondeterministic decisions, privilege abuse and unauthorized privilege propagation, context-related privacy leakage, and lack of trust in multi-agent collaboration. How to build a solid underlying security defense line and rebuild a full-stack trustworthy system while ensuring a high degree of agent autonomy has become the core bottleneck and key challenge constraining the evolution of agent technology toward large-scale industrial-grade deployment.","is_demo":0,"cover_url":"/api/forums/4215/cover?size=thumb","cover_alt":"Steering Agent Systems: Security Dilemmas and Trust Reconstruction of Agent OS","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4215","source_checked_at":"2026-10-30","agenda_json":"[[\"Analyzing Agent Security from a Data Perspective\",\"胡侠\"],[\"Agent Security Risks and Protection\",\"张源\"],[\"Toward Reliable Large Language Model Systems: Model, Software, and System Perspectives\",\"王帅\"],[\"Agent Operating Systems and Security\",\"夏虞斌\"],[\"AgentArmor – Agent Behavior Control\",\"吴烨\"],[\"Practice of Agent Security Solutions Based on Kunpeng Computing\",\"陈启军\"]]","code":"CNCC2026-4215","forum_type":"专题论坛","format":null,"theme_code":"AI安全与治理","theme_name":"AI安全与治理","venue":"西博城5号馆","room":"西博城5号馆-505","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":6,"forum_id":"CNCC2026-4215","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"驾驭智能体系统：Agent OS 的安全困境与信任重塑","original_topic":"AI安全与治理","original_description":"随着人工智能从单轮交互演进至具备自主决策、复杂规划及跨环境执行能力的智能体形态，智能体操作系统（Agent OS）作为新一代计算底座，正从根本上重构软件运行的基础逻辑。然而，智能体固有的自主性与高度动态性，已然消解了传统操作系统依托确定性代码与静态特权所建立的安全边界与隔离机制，由此引发多维系统级安全困境，包括非确定性决策失控、特权滥用与越权扩散、上下文相关的隐私泄露，以及多主体协同中的信任缺失等。如何在保障智能体高度自治的同时，构筑坚实底层安全防线并重建全栈可信体系，已成为制约智能体技术向大规模工业级部署演进的核心瓶颈与关键挑战。","people":[{"forum_id":4215,"role":"guest","name":"胡侠","affiliation":"上海人工智能实验室","sort_order":0},{"forum_id":4215,"role":"chair","name":"张殷乾","affiliation":"南方科技大学","sort_order":0},{"forum_id":4215,"role":"guest","name":"张源","affiliation":"复旦大学","sort_order":1},{"forum_id":4215,"role":"guest","name":"王帅","affiliation":"香港科技大学","sort_order":2},{"forum_id":4215,"role":"guest","name":"夏虞斌","affiliation":"上海交通大学","sort_order":3},{"forum_id":4215,"role":"guest","name":"吴烨","affiliation":"抖音集团","sort_order":4},{"forum_id":4215,"role":"guest","name":"陈启军","affiliation":"华为计算产品线","sort_order":5}],"host":"","image_url":"/api/forums/4215/cover"},{"id":4228,"title":"Breaking Barriers and Deepening Expertise: How Can Young Scholars Take Root and Break Through in the Intelligent Era?","topic":"General","kind":"专题论坛","chair":"文明","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-514","description":"This forum is initiated by the CCF 218 Club. The CCF 218 Club is mainly composed of scholars selected for the \"CCF Young Talent Development Program,\" which aims to support a group of young scholars who have already achieved relatively outstanding results and possess strong innovation capabilities; therefore, this forum has always focused on the growth and development of young scholars. As large models push research efficiency to new heights, young scholars no longer face the choice of \"whether to use AI\" but rather \"how to establish irreplaceable academic coordinates with AI empowerment.\" This forum takes \"Breaking Barriers and Deepening Expertise\" as its theme: \"Breaking Barriers\" refers to crossing disciplinary boundaries and seeking new growth points in interdisciplinary areas; \"Deepening Expertise\" emphasizes establishing a distinct identity in one's own field and avoiding being swept along by hot topics. The forum invites senior scholars to share how to anchor the main line of research amid change, and also invites young scholars to give presentations and engage in dialogue on topics such as academic choices, to help young scholars develop.","is_demo":0,"cover_url":"/api/forums/4228/cover?size=thumb","cover_alt":"Breaking Barriers and Deepening Expertise: How Can Young Scholars Take Root and Break Through in the Intelligent Era?","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4228","source_checked_at":"2026-10-30","agenda_json":"[[\"Finding the Right Coordinates: The Path of Software Quality Research in the Intelligent Wave\",\"谢晓园\"],[\"Creating \\\"Wireless Possibilities\\\": Ubiquitous Wireless Sensing\",\"张扶桑\"],[\"Rooted Breakthroughs: Efficient Compiler Optimization for Heterogeneous Hardware\",\"赵杰茹\"],[\"From Huawei to Renmin University: Industry-Problem-Driven Network Systems Research and Practice\",\"李彤\"],[\"System-Driven Research: From the Program Analysis Framework \\\"Tai'e\\\" to a Long-Term Research Roadmap\",\"谭添\"]]","code":"CNCC2026-4228","forum_type":"专题论坛","format":null,"theme_code":"综合","theme_name":"综合","venue":"西博城5号馆","room":"西博城5号馆-514","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4228","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"破壁与深耕——智能时代青年学者如何扎根和突围？","original_topic":"综合","original_description":"本论坛由CCF 218 Club发起。CCF 218 Club主要由入选“CCF青年人才发展计划”的学者组成，该计划旨在支持一批已经取得较优秀成绩、具有较强创新能力的青年学者，因此本论坛始终聚焦青年学者的成长与发展。当大模型将科研效率推向新高，青年学者面临的已非“用不用AI”的选择，而是“如何在AI赋能下建立不可替代的学术坐标”。本论坛以“破壁与深耕”为主题：“破壁”指跨越学科壁垒，在交叉地带寻找新生长点；“深耕”则强调在自身领域建立标识，避免被热点裹挟。论坛邀请资深学者分享如何在变动中锚定研究主线，也邀请青年学者围绕学术选择等议题进行汇报与对话，助力青年学着发展。","people":[{"forum_id":4228,"role":"chair","name":"文明","affiliation":"华中科技大学","sort_order":0},{"forum_id":4228,"role":"guest","name":"谢晓园","affiliation":"武汉大学","sort_order":0},{"forum_id":4228,"role":"guest","name":"张扶桑","affiliation":"北京航空航天大学","sort_order":1},{"forum_id":4228,"role":"guest","name":"赵杰茹","affiliation":"上海交通大学","sort_order":2},{"forum_id":4228,"role":"guest","name":"李彤","affiliation":"中国人民大学","sort_order":3},{"forum_id":4228,"role":"guest","name":"谭添","affiliation":"南京大学","sort_order":4}],"host":"","image_url":"/api/forums/4228/cover"},{"id":4239,"title":"Quantum Computing Industrialization: Sober Reflections Amid the Boom","topic":"Emerging Computing and Network Architectures","kind":"专题论坛","chair":"王俊超","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城4号馆-404","description":"Quantum computing is at a critical stage of moving from the laboratory to industrialization. In recent years, quantum computing companies at home and abroad have been active in financing, but there is a significant gap between technological maturity and commercial expectations. How to scientifically assess the current state of quantum computing technology, rationally view industry prospects and investment and financing conditions, and find a sustainable commercial closed loop and path to industrial implementation are core issues currently facing the industry.\nThis forum invites representatives from academia, the investment community, and enterprises at different links and different development stages of the quantum computing industry chain to jointly discuss feasible paths from technology to commerce for quantum computing from the perspectives of technological maturity, investment and financing status, full-machine commercialization, and supporting industry chains.","is_demo":0,"cover_url":"/api/forums/4239/cover?size=thumb","cover_alt":"Quantum Computing Industrialization: Sober Reflections Amid the Boom","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4239","source_checked_at":"2026-10-30","agenda_json":"[[\"Global Quantum Computing Industry Trends and Investment and Financing Landscape\",\"顾成建\"],[\"Product Practices and Application Exploration of Quantum Intelligent Software\",\"单小涵\"],[\"Domestic Quantum Computing Industry-Academia-Research Implementation: Closing the Loop from Technical Breakthroughs and Technology Transfer to Industry-Scenario Commercialization\",\"张辉\"],[\"Building a Local Quantum Industry Supporting Ecosystem for Quantum-Supercomputing-Intelligence Integration with Quantum Measurement and Control as the Core Pivot\",\"曾耿华\"],[\"Quantum Architecture Innovation Bridges Industry-Academia-Research Barriers and Accelerates the Industrialization of Quantum Hardware System Outcomes\",\"陈建鑫\"]]","code":"CNCC2026-4239","forum_type":"专题论坛","format":null,"theme_code":"新型计算与网络架构","theme_name":"新型计算与网络架构","venue":"西博城4号馆","room":"西博城4号馆-404","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":5,"forum_id":"CNCC2026-4239","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"量子计算产业化：热潮之下的冷思考","original_topic":"新型计算与网络架构","original_description":"量子计算正处于从实验室走向产业化的关键阶段。近年来国内外量子计算企业融资活跃，但技术成熟度与商业预期之间存在显著落差。如何科学评估量子计算的技术现状，理性看待产业前景与投融资情况，找到可持续的商业闭环与产业落地路径，是当前业界面临的核心问题。\n本论坛邀请学术界、投资界及量子计算产业链不同环节、不同发展阶段的企业代表，从技术成熟度、投融资现状、整机商业化、配套产业链等角度，共同研讨量子计算从技术到商业的可行路径。","people":[{"forum_id":4239,"role":"chair","name":"王俊超","affiliation":"复旦大学","sort_order":0},{"forum_id":4239,"role":"guest","name":"顾成建","affiliation":"北京鹬鸟科技有限公司","sort_order":0},{"forum_id":4239,"role":"guest","name":"单小涵","affiliation":"北京中科量枢科技有限公司","sort_order":1},{"forum_id":4239,"role":"guest","name":"张辉","affiliation":"本源量子计算科技（合肥）股份有限公司","sort_order":2},{"forum_id":4239,"role":"guest","name":"曾耿华","affiliation":"成都中微达信有限公司","sort_order":3},{"forum_id":4239,"role":"guest","name":"陈建鑫","affiliation":"清华大学","sort_order":4}],"host":"","image_url":"/api/forums/4239/cover"},{"id":4291,"title":"Coexistence, Safety, Governance, and Inclusiveness: The Proposition of Human-Machine Coexistence in the AI Era","topic":"AI Safety and Governance","kind":"专题论坛","chair":"吴飞","starts_at":"2026-10-24T13:30:00+08:00","ends_at":"2026-10-24T17:30:00+08:00","place":"西博城5号馆-511","description":"On July 17, 2026, General Secretary Xi Jinping delivered a keynote speech at the opening ceremony of the World Artificial Intelligence Conference and the High-Level Meeting on Global AI Governance, pointing out that humanity has to face the questions of the era: when machines begin to think, how should humans coexist with them? When algorithms participate in decision-making, how can safety be guaranteed? When technology challenges ethics, how can governance keep up? When the divide keeps widening, how can inclusiveness be achieved? These questions require in-depth reflection and joint answers from the international community. The forum will conduct discussions around the four dimensions of \"Coexistence, Safety, Governance, and Inclusiveness,\" inviting scholars from interdisciplinary fields such as AI ethics, public policy, and education to explore key issues including the boundaries and order of human-machine relations, the safety assurance of intelligent decision-making, the institutional supply of ethical governance, and the inclusive sharing of intelligent dividends. It strives to consolidate consensus, clarify differences, jointly promote research on AI ethics and governance, and help build a new order of human-machine coexistence that is people-centered, upward, and good-oriented.","is_demo":0,"cover_url":"/api/forums/4291/cover?size=thumb","cover_alt":"Coexistence, Safety, Governance, and Inclusiveness: The Proposition of Human-Machine Coexistence in the AI Era","source_url":"https://ccf.org.cn/CNCC2026/schedule_d_4291","source_checked_at":"2026-10-30","agenda_json":"[[\"Some Reflections on Computer Ethics Education in the Age of Artificial Intelligence\",\"郭耀\"],[\"Reflections on Framework Rules for Harmonious Human-Machine Development\",\"李超\"],[\"Privacy and Security of Generative Visual Models\",\"苗嘉旭\"],[\"Value Alignment: The Illusion of Consensus?\",\"林俊宇\"]]","code":"CNCC2026-4291","forum_type":"专题论坛","format":null,"theme_code":"AI安全与治理","theme_name":"AI安全与治理","venue":"西博城5号馆","room":"西博城5号馆-511","timezone":"Asia/Shanghai","is_assignment_forum":1,"missing_fields":"[]","notes":"[]","agenda_count":4,"forum_id":"CNCC2026-4291","lang":"en","tag":"en","i18n_tag":"en","translation_missing":false,"translation_source":"machine","original_title":"共处·安全·治理·普惠——AI时代的人机共生命题","original_topic":"AI安全与治理","original_description":"2026年7月17日，习近平总书记在世界人工智能大会暨人工智能全球治理高级别会议开幕式上的主旨讲话，指出人类不得不直面时代之问：当机器开始思考，人类如何与之相处？当算法参与决策，安全如何保障？当技术挑战伦理，治理如何跟上？当鸿沟不断拉大，普惠如何实现？这些问题需要国际社会深入思考、共同作答。论坛围绕\"共处·安全·治理·普惠\"四个维度展开研讨，邀请来自人工智能伦理、公共政策、教育等交叉领域的学者，探讨人机关系的边界与秩序、智能决策的安全保障、伦理治理的制度供给、智能红利的普惠共享等关键问题。力求凝练共识、厘清分歧，共同推动人工智能伦理与治理研究，助力构建以人为本、向上向善的人机共生新秩序。","people":[{"forum_id":4291,"role":"chair","name":"吴飞","affiliation":"浙江大学","sort_order":0},{"forum_id":4291,"role":"guest","name":"张圣宇","affiliation":"浙江大学","sort_order":0},{"forum_id":4291,"role":"guest","name":"郭耀","affiliation":"北京大学","sort_order":1},{"forum_id":4291,"role":"guest","name":"李超","affiliation":"上海交通大学","sort_order":2},{"forum_id":4291,"role":"guest","name":"苗嘉旭","affiliation":"哈尔滨工业大学（深圳）（哈尔滨工业大学深圳科技创新研究院）","sort_order":3},{"forum_id":4291,"role":"guest","name":"林俊宇","affiliation":"复旦大学","sort_order":4}],"host":"","image_url":"/api/forums/4291/cover"}],"lang":"en"}