arXiv cs.MA· Zekai Chen, Yihan Sun, Xunkai Li, Xun Wu, Xinmo Jin, Yinlin Zhu, Haodong Lu, Xuaner Chen, Rong-Hua Li·· 23 小时前AI 评分31
TRACE-MAS:诊断并修复多智能体系统中的协作故障
TRACE-MAS: Diagnosing and Repairing Collaboration Failures in Multi-Agent Systems
AI 导读
TRACE-MAS 是一个用于诊断和修复 LLM 多智能体系统协作故障的细粒度框架,通过学习良性轨迹中的正常协作模式,检测智能体状态变化、消息传递和任务依赖演化中的偏差。它通过来源感知归因定位责任执行元素,并选择最小干预恢复协作,在五个 MAS 基准上根因定位准确率提升 60.2%,任务恢复效果也优于现有强基线。
正文 · 原文
Abstract:LLM-based multi-agent systems (MAS) rely on not only individual agent behaviors but also the interactions that connect agents, including message exchanges and task dependencies. Existing MAS safeguards mainly perform agent-level anomaly detection, which becomes insufficient when failures originate from agent interactions and detected anomalies do not necessarily reveal their true sources. To address this limitation, we propose \textbf{TRACE-MAS}, a finer-grained framework for diagnosing and repairing MAS failures. TRACE-MAS learns normal collaborative execution patterns from benign trajectories and detects deviations in agent state changes, message transitions, and task dependency evolutions. It further identifies responsible execution elements through source-aware attribution and selects minimal interventions to restore collaboration while preserving task-critical capabilities. Experiments on five MAS benchmarks demonstrate that TRACE-MAS significantly improves root-cause localization accuracy by \textbf{60.2}\% and achieves consistent gains in task recovery over strong MAS safeguards.
| Subjects: | Machine Learning (cs.LG); Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2611.23362 [cs.LG] |
| (or arXiv:2611.23362v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2611.23362 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zekai Chen [view email]
[v1]
Tue, 17 Nov 2026 16:49:33 UTC (1,636 KB)
来源:arXiv cs.MA · arxiv.org