arXivDaily arXiv每日学术速递 周一至周五更新

期刊&会议

AAAI Conference on Artificial Intelligence · 会议 · Artificial Intelligence

共收录 9567
2510.27486 2026-04-21 cs.LG cs.AI

FedAdamW: A Communication-Efficient Optimizer with Convergence and Generalization Guarantees for Federated Large Models

FedAdamW: 一种具有收敛性和泛化保证的联邦大模型高效优化器

Junkang Liu, Fanhua Shang, Hongying Liu, Yuxuan Tian, Yuanyuan Liu, Jin Liu, Kewen Zhu, Zhouchen Lin

机构 * School of Computer Science and Technology, Tianjin University(天津大学计算机科学与技术学院) Medical College, Tianjin University(天津大学医学院) School of Artificial Intelligence, Xidian University(西安电子科技大学人工智能学院) State Key Lab of General Artificial Intelligence, School of Intelligence Science and Technology, Peking University(北京大学智能科学与技术学院通用人工智能国家重点实验室) Pazhou Laboratory (Huangpu), Guangzhou, Guangdong, China(广州黄埔 Pazhou 实验室)

AI总结 本文提出FedAdamW,通过局部修正机制和解耦权重衰减缓解联邦学习中的局部过拟合和方差问题,理论证明其收敛速率并实验证明其在语言和视觉Transformer模型中的有效性。

Journal ref AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.16266 2026-04-20 cs.CV

Hero-Mamba: Mamba-based Dual Domain Learning for Underwater Image Enhancement

Hero-Mamba:基于Mamba的双域学习用于水下图像增强

Tejeswar Pokuri, Shivarth Rai

机构 * Manipal Institute of Technology(曼普尔理工学院) Manipal Academy of Higher Education(曼普尔高等教育学院)

AI总结 本文提出Hero-Mamba,一种基于Mamba的网络,通过同时处理空间域和光谱域实现高效的水下图像增强,解决传统方法在长距离依赖建模和计算效率上的不足。

Comments Accepted at AI4ES Workshop AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2508.01345 2026-04-20 cs.CV

Predicting Video Slot Attention Queries from Random Slot-Feature Pairs

从随机槽-特征对预测视频槽注意力查询

Rongzhen Zhao, Jian Li, Juho Kannala, Joni Pajarinen

机构 * Department of Electrical Engineering and Automation, Aalto University(艾罗大学电气工程与自动化系) Gaoling School of Artificial Intelligence, Renmin University of China(中国人民大学研究生院人工智能学院) Department of Computer Science, Aalto University(艾罗大学计算机科学系) Center for Machine Vision and Signal Analysis, University of Oulu(奥卢大学机器视觉与信号分析中心)

AI总结 本文提出RandSF.Q方法,通过设计新的过渡器整合槽和特征信息,利用随机采样的槽-特征对预测查询,提升了视频对象中心学习的性能,显著优于现有方法。

Comments Accepted to AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.14832 2026-04-17 cs.SI cs.CY

Seeking Help, Facing Harm: Auditing TikTok's Mental Health Recommendations

寻求帮助,面对伤害:审计TikTok的心理健康推荐

Pooriya Jamie, Amir Ghasemian, Homa Hosseinmardi

AI总结 研究分析TikTok心理健康内容推荐机制,发现用户互动行为主导曝光结果,帮助导向搜索产生更多支持性内容,但有害内容仍存在。

Comments Accepted at the Proceedings of the International AAAI Conference on Web and Social Media (ICWSM 2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.13796 2026-04-16 cs.IR cs.LG

Driving Engagement in Daily Fantasy Sports with a Scalable and Urgency-Aware Ranking Engine

通过可扩展且具有紧迫感的排名引擎提升每日幻想体育的参与度

Unmesh Padalkar

机构 * Unmesh Padalkar

AI总结 本文提出一种可扩展的排名引擎,通过实时紧迫性特征和时间间隔编码提升每日幻想体育的推荐效果,实现9%的nDCG@1提升。

Journal ref Proceedings of the 40th AAAI Conference on Artificial Intelligence (AAAI-26), pp. 40378-40385, 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.24302 2026-04-16 cs.CR cs.CY

A Large-Scale Study of Telegram Bots

对Telegram机器人的大规模研究

Taro Tsuchiya, Haoxiang Yu, Tina Marjanov, Alice Hutchings, Nicolas Christin, Alejandro Cuevas

AI总结 本文首次大规模分析Telegram机器人,通过收集大量数据和开发自动交互系统,揭示了机器人在不同领域的应用及潜在的恶意用途。

Comments Proceedings of the 20th International AAAI Conference on Web and Social Media (ICWSM 2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.13739 2026-04-16 cs.LG stat.ML

Spectral Thompson sampling

谱泰普森采样

Tomas Kocak, Michal Valko, Remi Munos, Shipra Agrawal

机构 * SequeL team INRIA Lille - Nord Europe France(INRIA里尔-北欧洲法国SequeL团队) SequeL team INRIA Lille, France(INRIA里尔法国SequeL团队) Microsoft Research NE, USA(美国微软研究院NE部门) ML and Optimization Group Microsoft Research Bangalore, India(印度班加罗尔微软研究院机器学习与优化组)

AI总结 本文提出SpectralTS算法,通过图结构建模解决带状问题,利用有效维度d实现更高效的 regret 控制,适用于推荐系统和广告领域。

Comments Published at AAAI Conference on Artificial Intelligence (AAAI) 2014

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.13041 2026-04-16 cs.DB cs.AI

TableNet A Large-Scale Table Dataset with LLM-Powered Autonomous

TableNet:一个大规模表格数据集与LLM驱动的自主系统

Ruilin Zhang, Kai Yang

机构 * School of Computer Science and Technology, Tongji University(同济大学计算机科学与技术学院)

AI总结 TableNet通过LLM驱动的自主系统生成和识别表格,解决大规模高质量表格数据集的不足,提升表格结构识别的效率与精度。

Comments The 40th Annual AAAI Conference on Artificial Intelligence Bridge Program on Logic & AI

详情

展开后加载摘要…

URL PDF HTML 收藏
2602.13156 2026-04-16 cs.CR cs.AI

In-Context Autonomous Network Incident Response: An End-to-End Large Language Model Agent Approach

上下文自主网络事件响应:一种端到端的大语言模型代理方法

Yiran Gao, Kim Hammar, Tao Li

机构 * Department of Systems Engineering, City University of Hong Kong(香港城市大学系统工程系) Department of Electrical and Electronic Engineering, University of Melbourne(墨尔本大学电子与电气工程系)

AI总结 本文提出一种端到端的大语言模型代理方法,通过整合感知、推理、规划和行动功能,实现网络事件响应的自主学习与适应,实验表明其恢复效率比前沿模型快23%。

Comments 2026 AAAI Summer Symposium on Human-Aware AI Agents for the Cyber Battlefield

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.03523 2026-04-16 cs.AI cs.DS

Variance Computation for Weighted Model Counting with Knowledge Compilation Approach

基于知识编译方法的加权模型计数方差计算

Kengo Nakamura, Masaaki Nishino, Norihito Yasuda

机构 * Communication Science Laboratories, NTT, Inc.(NTT通信科学实验室)

AI总结 本文研究了加权模型计数方差的计算问题,提出了一种多项式时间算法,并证明了在不同知识编译格式下的计算复杂性,同时展示了其在贝叶斯网络不确定性分析中的应用。

Comments 25 pages; accepted for AAAI 2026 main track

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.12628 2026-04-15 math.OC cs.RO

A Comparison of Reinforcement Learning and Optimal Control Methods for Path Planning

强化学习与最优控制方法在路径规划中的比较

Qiang Le, Yaguang Yang, Isaac E. Weintraub

机构 * Department of Electrical and Computer Engineering, Hampton University(哈珀学院电气与计算机工程系) Air Warfare Directorate, Air Force Research Laboratory(空军研究实验室空战 Directorate)

AI总结 本文比较了强化学习与最优控制方法在路径规划中的应用,提出基于DDPG的方法能快速生成安全路径,但存在不可行区域,未来将改进奖励函数和探索其他方法。

Comments 8 pages, 9 figures, submitted to AAAI Conference

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.13789 2026-04-15 cs.CR cs.AI

Uncovering and Aligning Anomalous Attention Heads to Defend Against NLP Backdoor Attacks

揭示并对齐异常注意力头以防御NLP后门攻击

Haotian Jin, Yang Li, Haihui Fan, Lin Shen, Xiangfang Li, Bo Li

机构 * Institute of Information Engineering, Chinese Academy of Sciences(中国科学院信息工程研究所) State Key Laboratory of Cyberspace Security Defense(网络空间安全防御国家重点实验室) School of Cyber Security, University of Chinese Academy of Sciences(中国科学院大学网络安全学院)

AI总结 本文提出基于注意力相似性的后门检测方法,通过检测异常注意力头相似性来防御后门攻击,结合头部微调修复污染的注意力头,有效降低攻击成功率并保持下游任务性能。

Journal ref "Proceedings of the 40th AAAI Conference on Artificial Intelligence (AAAI 2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2405.18921 2026-04-15 cs.LG

GLANCE: Global Actions in a Nutshell for Counterfactual Explainability

GLANCE:为因果可解释性提供全球动作的简洁方法

Loukas Kavouras, Eleni Psaroudaki, Konstantinos Tsopelas, Dimitrios Rontogiannis, Nikolaos Theologitis, Dimitris Sacharidis, Giorgos Giannopoulos, Dimitrios Tomaras, Kleopatra Markou, Dimitrios Gunopulos, Dimitris Fotakis, Ioannis Emiris

机构 * Institute for the Management of Information Systems, Athena Research Center(信息管理系统研究所,雅典研究中心) Department of Electrical and Computer Engineering, National Technical University of Athens(电气与计算机工程系,雅典技术大学) Max Planck Institute for Software Systems(软件系统马克斯·普朗克研究所) Université Libre de Bruxelles(布鲁塞尔自由大学) FARI Institute, Belgium(比利时FARI研究所) Department of Informatics, Athens University of Economics and Business(信息系,雅典经济与商业大学) Department of Informatics and Telecommunications, National and Kapodistrian University of Athens(信息与电信系,雅典国家与卡波迪斯蒂亚大学) Archimedes, Athena Research Center(阿基米德,雅典研究中心)

AI总结 本文提出GLANCE算法,通过聚类方法平衡有效性、成本和动作数量,提升因果可解释性。

Journal ref 2026 Proceedings of the 40th Annual AAAI Conference on Artificial Intelligence

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.11833 2026-04-15 cs.LG

Uncertainty Quantification in CNN Through the Bootstrap of Convex Neural Networks

通过凸神经网络的自助法进行CNN的不确定性量化

Hongfei Du, Emre Barut, Fang Jin

机构 * The George Washington University(乔治华盛顿大学) Amazon.com, Inc.(亚马逊公司)

AI总结 本文提出基于凸神经网络的自助法框架,用于提升CNN预测不确定性的理论一致性,通过减少计算负载和引入迁移学习方法,在多个图像数据集上表现更优。

Comments 9 pages, 1 figure. Accepted at AAAI 2021

Journal ref Proceedings of the AAAI Conference on Artificial Intelligence, 35(13): 12078-12085, 2021

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.05914 2026-04-15 cs.CY

Designing Incident Reporting Systems for Harms from General-Purpose AI

为通用人工智能造成的伤害设计事件报告系统

Kevin Wei, Lennart Heim

AI总结 本文提出一个概念框架,探讨AI事件报告系统的制度设计,分析九个安全关键行业案例,提出美国AI事件报告的设计考量。

Comments Published in AAAI 2026. V2: Added Executive Summary, fixed hyperref line breaking issues

Journal ref Proceedings of the AAAI Conference on Artificial Intelligence. 40, 44 (Mar. 2026), 38016-38029

详情

展开后加载摘要…

URL PDF HTML 收藏
2501.16154 2026-04-15 cs.CL cs.AI

AdaMCoT: Rethinking Cross-Lingual Factual Reasoning through Adaptive Multilingual Chain-of-Thought

AdaMCoT:通过自适应多语言链式思考重新思考跨语言事实推理

Weihua Zheng, Xin Huang, Zhengyuan Liu, Tarun Kumar Vangani, Bowei Zou, Xiyan Tao, Yuhao Wu, Ai Ti Aw, Nancy F. Chen, Roy Ka-Wei Lee

AI总结 本文提出AdaMCOT框架,通过动态路由中间'思考语言'中的推理过程提升多语言事实推理能力,实验表明在低资源语言中表现显著提升。

Comments AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.05499 2026-04-14 cs.RO

TOSC: Task-Oriented Shape Completion for Open-World Dexterous Grasp Generation from Partial Point Clouds

TOSC:面向任务的形状补全用于从部分点云生成开放世界灵巧抓取

Weishang Wu, Yifei Shi, Zhiping Cai

AI总结 本文提出面向任务的形状补全方法,通过生成任务导向的接触区域完成方案,提升开放世界中灵巧抓取的性能,改进抓取位移和 Chamfer 距离。

Comments Accepted to AAAI 2026

Journal ref Proceedings of the AAAI Conference on Artificial Intelligence, 40(13), 10781-10789 (2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.11545 2026-04-14 cs.GT

Incremental Data-Driven Policy Synthesis via Game Abstractions

通过游戏抽象实现增量数据驱动的策略合成

Irmak Sağlam, Mahdi Nazeri, Alessandro Abate, Sadegh Soudjani, Anne-Kathrin Schmuck

AI总结 本文提出一种基于数据驱动和抽象的控制框架,通过增量游戏求解方法,在系统动态数据积累时逐步构建抽象游戏图、获胜区域和控制策略,实现对未知离散时间随机动态系统的控制策略合成。

Comments Presented at the 40th Annual AAAI Conference on Artificial Intelligence AAAI'26 (Oral)

详情

展开后加载摘要…

URL PDF HTML 收藏
2502.18026 2026-04-14 cs.LG cs.AI

ExPath: Targeted Pathway Inference for Biological Knowledge Bases via Graph Learning and Explanation

ExPath:通过图学习和解释进行生物知识库的靶向通路推断

Rikuto Kotoge, Ziwei Yang, Zheng Chen, Yushun Dong, Yasuko Matsubara, Jimeng Sun, Yasushi Sakurai

AI总结 本文提出ExPath框架,通过图学习和解释技术,整合实验数据推断生物网络中的靶向通路,实验表明其在Fidelity+和Fidelity-指标上优于基线方法。

Comments Accepted at AAAI 2026 (Main Technical Track)

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.09376 2026-04-14 cs.LG

From Decision Trees to Boolean Logic: A Fast and Unified SHAP Algorithm

从决策树到布尔逻辑:一种快速且统一的SHAP算法

Alexander Nadel, Ron Wettenstein

AI总结 本文提出WOODELF算法,结合决策树、博弈论和布尔逻辑,实现快速统一的SHAP计算,支持CPU和GPU高效运行,显著提升大规模数据处理速度。

Comments Published at AAAI 2026

Journal ref Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 40, No. 29, 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.07061 2026-04-14 cs.AI

Do LLMs Feel? Teaching Emotion Recognition with Prompts, Retrieval, and Curriculum Learning

LLMs能否感知情绪?通过提示、检索和课程学习进行情绪识别

Xinran Li, Yu Liu, Jiaqi Qiao, Xiujuan Xu

AI总结 本文提出PRC-Emo框架,结合提示工程、示范检索和课程学习,探索LLM在对话中感知情绪的能力,实验表明在IEMOCAP和MELD数据集上达到新的SOTA性能。

Comments Accepted at AAAI 2026

Journal ref Proc. AAAI Conf. on Artificial Intelligence, Vol. 40, No. 38, pp. 31778-31786 (2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.06443 2026-04-14 cs.LG

How Wide and How Deep? Mitigating Over-Squashing of GNNs via Channel Capacity Constrained Estimation

多宽多深?通过通道容量约束估计缓解图神经网络的过压缩

Zinuo You, Jin Zheng, John Cartlidge

AI总结 本文提出C3E框架,通过信息论将隐藏维度和深度的选择建模为非线性优化问题,缓解图神经网络的过压缩问题,提升表示学习性能。

Comments 29 pages, 11 figures. Author manuscript accepted for the 40th Annual AAAI Conference on Artificial Intelligence (AAAI-26), January 2026

Journal ref AAAI 2026, Proceedings of the AAAI Conference on Artificial Intelligence, 40(33), 27890-27898

详情

展开后加载摘要…

URL PDF HTML 收藏
2407.15389 2026-04-14 cs.LG cs.CR cs.DC

Poisoning with A Pill: Circumventing Detection in Federated Learning

用一粒药丸进行毒害:在联邦学习中规避检测

Hanxi Guo, Hao Wang, Tao Song, Tianhang Zheng, Yang Hua, Haibing Guan, Xiangyu Zhang

机构 * Purdue University(普渡大学)

AI总结 本文提出一种通用的攻击无关增强方法,通过在联邦学习训练中构造、生成并注入毒害(由现有攻击生成)到一个药丸(一种新型子网络结构)中,以提升现有毒害攻击的隐蔽性和有效性,揭示现有防御的不足。

Comments Accepted by AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.10775 2026-04-14 cs.CL cs.GT cs.LG

LLMs for Game Theory: Entropy-Guided In-Context Learning and Adaptive CoT Reasoning

大型语言模型用于博弈论:基于熵的上下文学习与自适应推理

Tommaso Felice Banfi, Sashenka Gamage

AI总结 本文提出一种基于LLM的框架,通过熵引导的链式推理和自适应上下文检索,提升离散博弈任务的推理能力,实验显示其显著提高决策质量。

Comments Published at the AAAI 2026 Bridge: Logical and Symbolic Reasoning in Language Models (OpenReview)

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.09282 2026-04-14 cs.SD cs.CL

End-to-end Contrastive Language-Speech Pretraining Model For Long-form Spoken Question Answering

端到端对比语言-语音预训练模型用于长形式语音问答

Jiliang Hu, Zuchao Li, Baoyuan Qi, Liu Guoming, Ping Wang

AI总结 本文提出CLSR模型,通过将音频特征转换为文本表示,提升长音频问答任务的性能,实验表明其优于现有方法。

Comments 12 pages, 7 figures, accepted by AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.00891 2026-04-14 cs.AI cs.CL

ChatCLIDS: Simulating Persuasive AI Dialogues to Promote Closed-Loop Insulin Adoption in Type 1 Diabetes Care

ChatCLIDS: 模拟说服性人工智能对话以促进1型糖尿病护理中的闭环胰岛素采用

Zonghai Yao, Talha Chafekar, Junda Wang, Shuo Han, Feiyun Ouyang, Junhui Qian, Lingxi Li, Hong Yu

AI总结 本文提出ChatCLIDS基准,用于评估基于LLM的说服对话在健康行为改变中的效果,通过模拟多轮交互和长期咨询场景,揭示当前LLM在行为改变中的局限性。

Comments Equal contribution for the first two authors. To appear in AAAI 2026 Special Track on AI for Social Impact

Journal ref AAAI 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2405.18374 2026-04-13 cs.CY cs.HC

Assessing How Hate, Counterspeech, and Toxicity Affect Hate Group Newcomers

评估仇恨、反仇恨言论和毒性如何影响仇恨群体新成员

Daniel Hickey, Matheus Schmitz, Daniel M. T. Fessler, Paul E. Smaldino, Kristina Lerman, Goran Murić, Keith Burghardt

AI总结 研究探讨了反仇恨言论对在线仇恨社区新成员参与的影响,发现反仇恨言论虽较仇恨言论毒性低,但可能加剧冲突并影响用户持续参与。

Comments 20 pages, 14 figures. arXiv admin note: text overlap with arXiv:2303.13641. Currently in press, Proceedings of the Twentieth International AAAI Conference on Web and Social Media (2024)

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.06838 2026-04-13 cs.CV

SparseCoop: Cooperative Perception with Kinematic-Grounded Queries

SparseCoop:基于运动学的协作感知

Jiahao Wang, Zhongwei Jiang, Wenchao Sun, Jiaru Zhong, Haibao Yu, Yuner Zhang, Chenyang Lu, Chuang Zhang, Lei He, Shaobing Xu, Jianqiang Wang

AI总结 本文提出SparseCoop,一种完全稀疏的协作感知框架,用于3D检测与跟踪,通过运动学实例查询、粗到细聚合模块和协作实例去噪任务,实现高效且鲁棒的协作感知。

Comments Accepted by AAAI 2026

Journal ref Proceedings of the AAAI Conference on Artificial Intelligence, vol. 40, no. 12, pp. 9876-9884 (2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.06756 2026-04-13 cs.LG

Dual Mamba for Node-Specific Representation Learning: Tackling Over-Smoothing with Selective State Space Modeling

双Mamba用于节点特定表示学习:通过选择性状态空间建模解决过平滑问题

Xin He, Yili Wang, Yiwei Dai, Xin Wang

AI总结 本文提出DMbaGCN框架,通过整合Mamba模型从局部和全局视角缓解深度图神经网络中的过平滑问题,采用LSEMba和GCAMba模块提升节点可区分性。

Comments Accepted by The 40th Annual AAAI Conference on Artificial Intelligence (AAAI 2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2503.06983 2026-04-13 cs.CV cs.RO

Griffin: Aerial-Ground Cooperative Detection and Tracking Dataset and Benchmark

Griffin:空地协同检测与跟踪数据集与基准

Jiahao Wang, Xiangyu Cao, Jiaru Zhong, Yuner Zhang, Zeyu Han, Haibao Yu, Chuang Zhang, Lei He, Shaobing Xu, Jianqiang Wang

AI总结 本文提出Griffin数据集,包含250个动态场景,用于评估空地协同检测与跟踪的性能,提供统一的基准框架以评估通信效率和鲁棒性。

Comments Accepted by AAAI 2026

Journal ref Proceedings of the AAAI Conference on Artificial Intelligence, vol. 40, no. 12, pp. 9867-9875 (2026)

详情

展开后加载摘要…

URL PDF HTML 收藏