paperbot · PL 论文追踪

全部论文 待读 收藏

P2 Efficient Failure Management for Multi-Agent Systems with Reasoning Trace Representation

基于推理轨迹表示的多智能体系统高效故障管理
FSE Proceedings of the 34th ACM International Conference on the Foundations of Software Engineering2026匹配度 3/10 · 论文研究多智能体 LLM 系统的故障检测与缓解，与 Agentic 程序可靠性有一定关联，但它依赖无监督对比学习做推理轨迹表示，并不涉及静态分析、形式化验证、浮点程序分析或贝叶斯程序分析，也未将 LLM/Agent 用作改进程序分析的工具。整体更偏向 LLM 系统运维与测试，而非用户核心方向。引用 6 inbox
Lingzhe Zhang, Tong Jia, Mingyu Wang, Weijie Hong, Chiming Duan, Minghua He, Rongqian Wang, Xi Peng, Meiling Wang, Nicholas Zhang, Renhai Chen, Ying Li
机器学习与程序语言测试、调试与修复
crossref: 2026-09-12dblp: 2026-09-12arxiv: 2026-09-12

AI 速览

一句话：提出 EAGER 框架，用无监督推理轨迹对比学习编码智能体内部推理与智能体间协作，实现多智能体系统的实时故障检测、诊断与缓解。

基于大语言模型的多智能体系统（MAS）正成为软件系统设计的新范式，其可靠性与可用性依赖于有效的故障管理。现有方法多采用逐条轨迹的推理方式，效率低下，且忽视历史故障模式，限制了诊断准确性。本文首先开展初步实证研究，论证利用历史故障模式增强 MAS 故障管理的必要性、潜力与挑战。在此基础上提出 EAGER：一种基于推理轨迹表示的高效故障管理框架。EAGER 采用无监督的推理范围对比学习（reasoning-scoped contrastive learning），同时编码智能体内部的推理过程与智能体间的协作行为，从而在历史故障知识引导下支持实时的步骤级故障检测、诊断与自反缓解。在三个开源 MAS 上的初步评估验证了 EAGER 的有效性，并指出了面向可靠多智能体系统运维的未来研究方向。

问题
基于 LLM 的多智能体系统日益复杂自治，但现有故障管理依赖逐条轨迹推理、效率低，且未利用历史故障模式，导致诊断准确性受限。
方法
先通过实证研究论证历史故障模式的价值，再提出 EAGER：使用无监督的推理范围对比学习，将智能体内部推理与智能体间协作编码为推理轨迹表示，并借助历史故障知识实现步骤级故障检测、诊断与自反缓解。
评估
在三个开源多智能体系统上进行初步评估，验证了 EAGER 的有效性，并讨论了未来可靠 MAS 运维的研究方向。
原文摘要（Abstract）

Large Language Models (LLM)-based Multi-Agent Systems (MASs) have emerged as a new paradigm in software system design, increasingly demonstrating strong reasoning and collaboration capabilities. As these systems become more complex and autonomous, effective failure management is essential to ensure reliability and availability. However, existing approaches often rely on per-trace reasoning, which leads to low efficiency, and neglect historical failure patterns, limiting diagnostic accuracy. In this paper, we conduct a preliminary empirical study to demonstrate the necessity, potential, and challenges of leveraging historical failure patterns to enhance failure management in MASs. Building on this insight, we propose EAGER, an efficient failure management framework for multi-agent systems based on reasoning trace representation. EAGER employs unsupervised reasoning-scoped contrastive learning to encode both intra-agent reasoning and inter-agent coordination, enabling realtime step-wise failure detection, diagnosis, and reflexive mitigation guided by historical failure knowledge. Preliminary evaluations on three open-source MASs demonstrate the effectiveness of EAGER and highlight promising directions for future research in reliable multi-agent system operations.

链接与引用

DOI 原文 · arXiv · PDF（开放获取） · DBLP

BibTeX 
@inproceedings{abs-2603-21522,
  title = {Efficient Failure Management for Multi-Agent Systems with Reasoning Trace Representation},
  author = {Lingzhe Zhang and Tong Jia and Mingyu Wang and Weijie Hong and Chiming Duan and Minghua He and Rongqian Wang and Xi Peng and Meiling Wang and Nicholas Zhang and Renhai Chen and Ying Li},
  booktitle = {FSE},
  year = {2026},
  doi = {10.1145/3803437.3805560}
}