SearchSwarm: Towards Delegation Intelligence in Agentic LLMs for Long-Horizon Deep Research
Tsinghua University(清华大学);Peking University(北京大学);Ant Group(蚂蚁集团);Gaoling School of Artificial Intelligence, Renmin University of China(中国人民大学高瓴人工智能学院)
Xiaochong Lan, Quan Chen, Kun Tao, Xinyu Tang, Tianshu Wang, Qianggang Cao, Xinyu Kong, Zujie Wen, Zhiqiang Zhang, Jun Zhou
Large language models are increasingly expected to handle complex, long-horizon real-world tasks whose context demands can grow without bound, yet model context windows remain inherently finite. Recent work explores a paradigm where a main agent decomposes tasks and dispatches subtasks to subagents, which execute and return only summarized results, conserving the main agent's context budget. However, performing this well requires delegation intelligence: the ability to decompose complex tasks, determine when and what to delegate, and integrate returned results into the ongoing workflow. Training data for this capability is scarce in naturally occurring text, and to our knowledge, how to synthesize such data and train models to acquire this capability remains largely unexplored in the open-source community. To bridge this gap, we present a preliminary exploration targeting deep research, a representative long-horizon agent task. Specifically, we design a harness that guides the model toward high-quality task decomposition and delegation, while constraining subagents to return results properly to support the main agent's workflow. The harness-guided trajectories naturally encode correct delegation decisions, which we use as supervised fine-tuning data to internalize delegation intelligence into model weights. Our resulting model, SearchSwarm-30B-A3B, achieves 68.1 on BrowseComp and 73.3 on BrowseComp-ZH, the best results among all models of comparable scale. We will release our harness, model weights, and training data to facilitate future research.
Agentic programs: an emerging form of scientific software in computational materials science
Research Institute of Advanced Materials, Seoul National University(首尔国立大学先进材料研究所);Department of Materials Science and Engineering, Seoul National University(首尔国立大学材料科学与工程系);Center for AI and Natural Sciences, Korea Institute for Advanced Study(韩国高等研究院人工智能与自然科学中心)
Distributed Legal Infrastructure for a Trustworthy Agentic Web
Institute for Technoscience and Society(科技与社会研究所);Berkeley Center for Law & Technology(伯克利法与科技中心);U.C. Berkeley Law School(伯克利法学院);Technical University of Munich(慕尼黑技术大学);Università degli Studi dell’Aquila(阿奎拉大学);University of Oxford(牛津大学);Existential Risk Alliance(存在风险联盟);NYU Shanghai(纽约大学上海分校);Sun Yat-sen University(中山大学);Shanghai Innovation Institute(上海创新研究院);Norwegian University of Science and Technology(挪威科学技术大学)
Oxford Martin AI Governance Initiative(牛津马丁人工智能治理研究所);Access Partnership;University of Oxford(牛津大学);Google DeepMind(谷歌DeepMind);Cooperative AI Foundation(协作人工智能基金会)
MIT(麻省理工学院);Harvard Law School(哈佛法学院);University of California, Berkeley(加州大学伯克利分校);University of Oxford(牛津大学);Centre for the Governance of AI(AI治理中心);Stanford University(斯坦福大学)
针对自主 AI 智能体的授权与问责问题,论文提出基于 OAuth 2.0 和 OpenID Connect 扩展的可认证、可审计委托框架,并将自然语言权限转为访问控制策略。
A Practical Guide to Agentic AI Transition in Organizations
Old Dominion University(旧 Dominion 大学);Florida International University(佛罗里达国际大学);Nanyang Technological University(南洋理工大学);University of Colombo(科伦坡大学);Center for Wireless Communications, University of Oulu(无线通信中心,奥卢大学);University of Sri Jayewardenepura(斯里贾yenepura大学);Accenture Technology Labs(埃森哲技术实验室)
Agentivism: a learning theory for the age of artificial intelligence
School of Education, Tsinghua University(清华大学教育学院);Faculty of Education and School of Computing & Data Science, The University of Hong Kong(香港大学教育学院及计算与数据科学学院);Faculty of Information Technology, Monash University(墨尔本大学信息技术学院)
University of Washington(华盛顿大学);University of California, Santa Barbara(加州大学圣芭芭拉分校);Stanford University(斯坦福大学);Carnegie Mellon University(卡内基梅隆大学);Northwestern University(西北大学);University of Notre Dame(圣母大学);University of California, Berkeley(加州大学伯克利分校);Michigan State University(密歇根州立大学);MIT-IBM Watson AI Lab(MIT-IBM沃森人工智能实验室);Bake AI;King Abdulaziz City for Science and Technology(国王阿卜杜勒阿齐兹科技城);Western Washington University(西雅图华盛顿大学);University of Chicago(芝加哥大学)
From Assistance to Agency: Rethinking Autonomy and Control in CI/CD Pipelines
Faculty of Information University of Toronto Toronto Ontario Canada(信息学院多伦多大学多伦多安大略加拿大);Department of Computer Science Trent University Peterborough Ontario Canada(计算机科学系特伦特大学彼得伯格安大略加拿大);University of Toronto(多伦多大学);Trent University(特伦特大学)