机制可解释性能否解释 LLM 智能体失败?网页元素幻觉案例研究
Can Mechanistic Interpretability Help Explain LLM Agent Failures? A Case Study on Web Element Hallucination
AI 导读
一篇 arXiv 论文考察机制可解释性(MI)工具在解释 LLM 智能体失败时的局限,以 Qwen2.5-7B-Instruct 与真实网站交互时的网页元素幻觉为案例。
正文 · 原文
Abstract:Mechanistic Interpretability (MI) has made significant progress in explaining the internal computations of Large Language Models (LLMs); however, these advances have yet to demonstrate comparable success in explaining the decision-making processes of LLM-based agents. This paper examines the challenges MI tools face when applied to realistic agentic failures. We focus on one such failure: web agents hallucinating actions that are not available on a website. We apply a series of mechanistic analyzes to Qwen2.5-7B-Instruct as it interacts with live websites. We find that a linear probe trained on the last prompt token can predict hallucinations before generation begins (0.80 AUC) and steering along a probe direction corrects 85\% of hallucinations, but disrupts 7\% of valid actions. However, these results only provide a partial explanation of why the agent hallucinates. The probe direction is sensitive to prompt formatting, losing much of its steering effect after a minor change to how actions are represented. Additionally, circuit identification can be difficult as the computation is distributed across multiple components and positions, and the information the agent relies on varies across examples. We discuss how these observations challenge the assumptions underlying standard MI methods for circuit discovery and feature identification, and outline future directions for making these methods better suited to explain failures in LLM-based agents.
| Subjects: | Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.18514 [cs.AI] |
| (or arXiv:2611.18514v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.18514 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Mohamed Aghzal [view email]
[v1]
Fri, 13 Nov 2026 18:30:58 UTC (1,328 KB)
来源:arXiv cs.AI · arxiv.org