arXiv：cs.AI· Ayse Sila Okcu, Weisheng Wang, Tongmeng Xie, Niall Dalton, Angelos Nalmpantis, Raghu Arghal, Mario Giulianelli·· 20 小时前
语言模型智能体中的信念-行动鸿沟
The belief-action gap in language model agents
AI 导读
研究用报告式与表征式探针估计任务相关信念、用逆强化学习估计行动价值，发现语言模型智能体在文本网格世界中往往能可靠地表征状态与价值信息，却未在行为中体现。追踪推理过程中的行动偏好显示，决策会在某个承诺点附近收敛，报告的不确定性也随之下降。强制行动实验表明，来自其他决策的推理前缀可被移植，而推理窗口激活编辑对行动的控制有限。
正文
Abstract:Language model (LM) agents can fail for different reasons, requiring different explanations and interventions. We study how state beliefs, action values, and action preferences interact as LM agents reason and commit to decisions in text-rendered grid worlds. We estimate task-relevant beliefs using report-based and representation-based probes, and action values using inverse reinforcement learning. We find that task-relevant state and value information is often represented more reliably than it is reflected in behaviour. Tracking action preferences throughout reasoning further shows that decisions sharpen around a commitment point and reported belief uncertainty decreases near commitment, although individual belief changes do not consistently explain shifts in action preference. Finally, forced actions follow reasoning prefixes from other decisions when transplanted, whereas the tested reasoning-window activation edits provide limited control over the action. Together, these results suggest that some failures arise not from missing task-relevant information, but from how available information is translated into action. Our measurements distinguish information availability from belief-action consistency without by themselves identifying the cause of each failure. Our code is available at: this https URL
| Comments: | 25 pages, 10 figures; includes appendices |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.17444 [cs.LG] |
| (or arXiv:2610.17444v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.17444 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Weisheng Wang [view email]
[v1]
Mon, 12 Oct 2026 22:04:36 UTC (1,146 KB)
来源：arXiv：cs.AI · arxiv.org