走错路并不会毁掉旅程：面向LLM智能体的偏差引导技能自进化
A Wrong Turn Does Not Ruin the Journey: Deviation-Guided Skill Self-Evolution for LLM Agents
查看机构详情
- University of Chinese Academy of Sciences（中国科学院大学）
- University of Science and Technology of China（中国科学技术大学）
- Meituan（美团）
机构由 AI 辅助整理，请以论文原文为准。
浏览论文内容
中文总结 AI 辅助
针对LLM智能体技能自进化，提出偏差点引导框架SkillPivot，识别失败轨迹中的有效前缀与错误后缀，通过教师对比生成局部技能更新，在多个基准上优于现有方法。
中文摘要 AI 辅助
大型语言模型智能体日益依赖自然语言技能来解决复杂的工具使用任务。然而，此类任务通常允许多种有效的解决路径，因此通过强制失败的轨迹匹配固定的成功轨迹来改进技能是不合适的。此外，失败的轨迹很少是完全错误的：智能体可能首先收集有用的证据并取得有意义的进展，但随后偏离到错误的后续部分。因此，我们认为技能自进化应识别生产性问题解决开始失效的位置，而不是对整个失败进行粗略反思。基于这一见解，我们提出了SkillPivot，一种以偏差点引导的技能自进化框架。SkillPivot通过执行有效性、目标进展和动作多样性来检测从有用前缀到错误后缀的转变。随后，一个更强的教师模型从相同前缀继续，并在相同的交互历史下产生成功的替代方案。通过对比学生失败的后缀与教师成功的后缀，SkillPivot生成局部技能更新，同时保留已有有效的指导。在ToolQA、LogicBench和WildClawBench上的实验表明，SkillPivot持续优于竞争性的技能进化方法，改进了多个智能体模型，并产生了紧凑、可迁移的技能更新。
英文摘要
Large language model agents increasingly rely on natural-language skills to solve complex tool-use tasks. However, such tasks often admit multiple valid solution paths, making it inappropriate to improve skills by forcing failed trajectories to match a fixed successful trajectory. Moreover, failed trajectories are rarely entirely wrong: an agent may first collect useful evidence and make meaningful progress, but later deviate into an erroneous suffix. We therefore argue that skill self-evolution should identify where productive problem solving begins to break down, rather than reflect coarsely over the entire failure. Based on this insight, we propose SkillPivot, a deviation-point-guided framework for skill self-evolution. SkillPivot detects the transition from a useful prefix to an erroneous suffix using execution validity, goal progress, and action diversity. A stronger teacher then continues from the same prefix and produces a successful alternative under the same interaction history. By contrasting the student's failed suffix with the teacher's successful suffix, SkillPivot generates localized skill updates while preserving already effective guidance. Experiments on ToolQA, LogicBench, and WildClawBench show that SkillPivot consistently outperforms competing skill-evolution methods, improves multiple agent models, and produces compact, transferable skill updates.
