arXiv：cs.AI· Jingyu Liu, Zhiwen Wang, Yuxin Jing, Huanyu Zhou, Yong Liu·· 4 小时前AI 评分44
语言智能体为何无法将历史转化为经验：动作校准研究
When History Fails to Become Experience: Action Calibration in Language Agents
AI 导读
研究发现语言智能体并未可靠地将过去动作与其结果关联，即使打乱历史动作顺序，任务成功率也仅小幅下降。将每个返回的观测显式标注为前一动作的结果，可在不引入新环境信息的情况下提升任务成功率并减少动作重复。基于此，研究者提出一种学习型校准器，通过重新评估过去动作并选择性记录经验来指导后续决策，效果优于单纯的结果标注。
正文
Abstract:Language agents should draw on prior attempts and environmental feedback to improve subsequent decisions within the same task. However, providing additional interaction history can sometimes reduce task success, suggesting that agents do not consistently use this information effectively. To investigate this limitation, we examine how agents use history. We find that history improves task completion overall, yet much of this benefit persists even when past actions are shuffled. Disrupting the correspondence between actions and observations causes only a modest decline in task success. We therefore hypothesize that agents do not reliably connect past actions with their outcomes when deciding how to proceed. To test this hypothesis, we explicitly label each returned observation as the outcome of the preceding action. This simple annotation improves task success and reduces next-action repetition without introducing new environmental information. Building on this insight, we introduce a learned calibrator that explicitly reassesses past actions and selectively records experience to guide subsequent decisions, improving task success beyond outcome labeling alone.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.02769 [cs.CL] |
| (or arXiv:2610.02769v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.02769 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jingyu Liu [view email]
[v1]
Fri, 2 Oct 2026 03:50:34 UTC (207 KB)
来源：arXiv：cs.AI · arxiv.org