From Flat Logs to Causal Graphs: Hierarchical Failure Attribution for LLM-based Multi-Agent Systems

TL;DR

This paper introduces CHIEF, a hierarchical causal graph framework for better failure attribution in LLM-based multi-agent systems, improving accuracy over existing flat log analysis methods.

Contribution

It presents a novel hierarchical causal modeling approach with oracle-guided backtracking and counterfactual attribution for more precise failure analysis in MAS.

Findings

CHIEF outperforms eight state-of-the-art baselines on the Who&When benchmark.

Hierarchical causal modeling improves failure attribution accuracy.

Ablation studies confirm the effectiveness of each module.

Abstract

LLM-powered Multi-Agent Systems (MAS) have demonstrated remarkable capabilities in complex domains but suffer from inherent fragility and opaque failure mechanisms. Existing failure attribution methods, whether relying on direct prompting, costly replays, or supervised fine-tuning, typically treat execution logs as flat sequences. This linear perspective fails to disentangle the intricate causal links inherent to MAS, leading to weak observability and ambiguous responsibility boundaries. To address these challenges, we propose CHIEF, a novel framework that transforms chaotic trajectories into a structured hierarchical causal graph. It then employs hierarchical oracle-guided backtracking to efficiently prune the search space via sybthesized virtual oracles. Finally, it implements counterfactual attribution via a progressive causal screening strategy to rigorously distinguish true root causes from propagated symptoms. Experiments on Who&When benchmark show that CHIEF outperforms eight strong and state-of-the-art baselines on both agent- and step-level accuracy. Ablation studies further confirm the critical role of each proposed module.