arXiv cs.AI· Zijian Jiang, Xilu Wang·· 13 小时前AI 评分32
工具智能体通过行为定位实现鲁棒软技能适配
Robust Soft Skill Adaptation via Behavioral Localization in Tool-Using Agents
AI 导读
研究提出 Behavioral Localization 方法，通过对比冻结智能体有无文本技能时的表现，定位技能真正起作用的轨迹位置，并用这些位置监督短软提示。
正文 · 原文
Abstract:Text skills can extend the capabilities of tool-using agents, but long skill descriptions increase computational costs. Soft prompts provide a compact alternative, yet learning from entire successful trajectories may also introduce unnecessary behavioral changes. We find that the useful effect of a text skill is highly concentrated in a small subset of trajectory positions. Based on this observation, we propose Behavioral Localization, which compares the frozen agent with and without the text skill to identify positions where the skill both increases the likelihood of successful tokens and meaningfully changes the next token distribution. We use these localized positions to supervise a short soft prompt. Experiments on SpreadsheetBench with a frozen Qwen3.6-35B-A3B agent show that the top 10\% of trajectory positions capture 96.12\% of the total positive skill gain. In the shared prompt setting, selective supervision improves success from 30.36\% with full trajectory to 36.79\%. Under task-specific conditions, using the top 10\% localized positions achieves 57.38\%, outperforming both the full text skill and the on-policy OPCD baseline. These results show that a small set of behaviorally relevant trajectory positions can provide effective supervision of soft skills.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.30041 [cs.AI] |
| (or arXiv:2610.30041v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.30041 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zijian Jiang [view email]
[v1]
Wed, 21 Oct 2026 17:02:06 UTC (723 KB)
来源：arXiv cs.AI · arxiv.org