发表机构
Computer Network Information Center, Chinese Academy of Sciences; Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算机网络信息中心; 中国科学院计算技术研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对多智能体系统故障归因的现有方法存在不足，该研究提出即插即用的DUOTRACE，采用先检测后归因范式，提升了智能体级和步骤级归因准确率。
AI 中文摘要
基于大语言模型（LLM）的智能体通过多步推理展现出解决复杂任务的强大潜力，但仍易受执行故障影响，因此准确的故障归因对提升智能体可靠性至关重要。现有基于拓扑和频谱的方法利用轨迹结构却常忽略细粒度语义，而基于LLM的归因方法虽能捕捉语义线索，却在长轨迹上存在长上下文退化问题。为应对这些挑战，我们提出DUOTRACE，一种适用于LLM故障归因的即插即用检测过滤器。DUOTRACE遵循“先检测后归因”范式：首先检测异常执行，再为下游基于LLM的归因方法提供聚焦的轨迹证据。为实现对智能体轨迹的基于变分自编码器（VAE）的有效异常检测，DUOTRACE整合了双视图语义-结构节点表示、基于树长短期记忆网络（Tree-LSTM）的轨迹编码器，以及基于前缀链和LLM的数据增强，以处理异构节点、分层执行结构和有限的故障数据。对6种基于LLM的归因基线的实验表明，DUOTRACE分别将智能体级和步骤级归因准确率提升了8.7%和7.0%。
英文摘要
Large language model (LLM)-based agents have shown strong potential in solving complex tasks through multi-step reasoning, yet they remain vulnerable to execution failures. Accurate failure attribution is therefore critical for improving agent reliability. Existing topology- and spectrum-based methods exploit trajectory structures but often overlook fine-grained semantics, while LLM-based attribution methods capture semantic cues but suffer from long-context degradation over lengthy trajectories. To address these challenges, we propose DUOTRACE, a plug-and-play detection filter for LLM-based failure attribution. DUOTRACE follows a detect-before-attribute paradigm: it first detects anomalous executions and then supplies focused trajectory evidence to downstream LLM-based attribution methods. For effective VAE-based anomaly detection on agent trajectories, DUOTRACE integrates dual-view semantic-structural node representations, a Tree-LSTM-based trajectory encoder, and prefix-chain- and LLM-based data augmentation to handle heterogeneous nodes, hierarchical execution structures, and limited failure data. Experiments with six LLM-based attribution baselines show that DUOTRACE improves agent-level and step-level attribution accuracy by 8.7% and 7.0%, respectively.
Comments17 pages, 5 figures, 13 tables
