arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2610.19342cs.MAcs.AI
收藏

智能体团队为何跨边界失败?具有私有状态的智能体间系统的失败分类

Why Do Agent Teams Fail Across Boundaries? A Failure Taxonomy for Agent-to-Agent Systems with Private States

Salesforce;University of North Carolina(北卡罗来纳大学);Singapore University of Technology and Design(新加坡科技设计大学);Nanyang Technological University(南洋理工大学);University of British Columbia(不列颠哥伦比亚大学)
Sarath Shekkizhar, Prathyusha Jwalapuram, Zixuan Ke, Vaidehi Patil, Jiayu Wang, Srijan Bansal, Trong Huy Nguyen, Dingfeng Jiang, Dongqi Liu, Ruochen Zhao, Wei L… 展开作者

Sarath Shekkizhar, Prathyusha Jwalapuram, Zixuan Ke, Vaidehi Patil, Jiayu Wang, Srijan Bansal, Trong Huy Nguyen, Dingfeng Jiang, Dongqi Liu, Ruochen Zhao, Wei Lu, Giuseppe Carenini, Mohit Bansal, Silvio Savarese, Shafiq Joty

AI总结:

该研究针对具有私有状态的智能体间系统,分析2500余次运行数据,提出含20种失败模式的分类$A^2$-Failures,发现三类问题并主张智能体需按自身目标单独评分。

摘要:

智能体正越来越多地与代表不同委托方(它们所服务的个人或组织)的其他智能体交互,例如买方的个人智能体与企业的卖方智能体进行谈判。在这类智能体间(Agent-to-Agent,A2A)场景中,每个智能体都有自己的环境,且持有其他智能体无法看到的信息和权限。此外,每个智能体的利益可能是一致的(合作型)、部分一致的(混合动机型)或对立的(竞争型)。现有的多智能体系统失败研究仅假设存在一个编排器、共享工作空间或委托方,我们研究当这些假设被移除时A2A系统如何失败。我们研究了来自不同智能体框架和模型构建的A2A系统的2500多次记录运行,涵盖了8个领域(法律、客户关系管理、安全、商业等)的488种场景。从这些轨迹中,我们推导了$A^2$-Failures,这是一个分为四类的20种A2A失败模式的分类:(i)发现与委托,(ii)基础与记录一致性,(iii)验证,以及(iv)行为与保真度。在研究的框架、模型和场景中,有三个观察结果成立:(1)智能体将任务与提示中提到的对等体名称匹配,而非主动搜索可用对等体;(2)智能体不会主动对彼此进行事实核查;(3)仅对最终输出评分会奖励主导智能体,同时隐藏内部协调、验证和行为故障。我们的研究主张进行全面评估,除了任务结果外,还应根据智能体自身的目标对其单独评分。我们在https://huggingface.co/datasets/a2a-salesforce/a2-failures发布了该分类背后的运行日志。

英文摘要:

Agents are increasingly interacting with other agents that act for different principals (people or organizations on whose behalf they work), such as a buyer's personal agent negotiating with an enterprise's seller agent. In such agent-to-agent (A2A) scenarios, each agent has its own environment and holds information and authority that the other cannot see. Moreover, the interests of each agent may be aligned (cooperative), partly aligned (mixed-motive), or opposed (competitive). Existing multi-agent system failure studies assume one orchestrator, shared workspace, or single principal. We ask how A2A systems fail when these assumptions are removed. We study more than 2,500 recorded runs of A2A systems built from different agent harnesses and models, spanning 488 scenarios across eight families (legal, CRM, security, commerce, etc.). From these traces, we derive $A^2$-Failures, a taxonomy of 20 A2A failure modes in four categories: (i) discovery and delegation, (ii) grounding and record consistency, (iii) verification, and (iv) conduct and fidelity. Three findings hold across every harness, model, and scenario studied: (1) discovery and interaction are driven by the phrasing of the task rather than the availability of peer agents; (2) no agent verifies a peer's factual claim on its own initiative, so an A2A system inherits the reliability of its least reliable participant unless the protocol or mechanism performs the check; and (3) the aggregate outcome score credits the agent completing the task while staying blind to the coordination, verification, and conduct failures beneath it, so A2A systems must be evaluated per agent, against each agent's own principal. We release the run logs behind the taxonomy at https://huggingface.co/datasets/a2a-salesforce/a2-failures.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑