arXiv cs.AI· Yujun Wang, Jiqiang Li, Yinuo Lu, Jinhe Bi, Aniri, Wenxuan Ye, Boliang Liu, Sikuan Yan, Shuning Wang, Xuebing Zhou, S{\"o}ren Pirk, Hinrich Sch{\"u}tze, Yunpu Ma·· 2 小时前AI 评分42
EvoSkill-S1:用遗传启发算子进化可复用技能
EvoSkill-S1: Evolving Reusable Skills with Genetic-Inspired Operators
AI 导读
Yujun Wang 等研究者提出 EvoSkill-S1,一个以强化学习直接进化智能体可复用技能的框架。它通过修订、交叉修订、变异、双重反思和杂交五种遗传启发算子,从困难负样本失败中生成子技能,并用下游任务奖励训练技能进化器、以评估器引导回写保留有益更新。
正文 · 原文
Authors:Yujun Wang, Jiqiang Li, Yinuo Lu, Jinhe Bi, Aniri, Wenxuan Ye, Boliang Liu, Sikuan Yan, Shuning Wang, Xuebing Zhou, S{ö}ren Pirk, Hinrich Sch{ü}tze, Yunpu Ma
Abstract:Reusable skills allow language-model agents to accumulate and reuse problem-solving experience. However, most existing methods improve a skill bank only indirectly through rollout-driven accumulation and heuristic reflection, often resulting in shallow and weakly controlled updates. We present EvoSkill-S1, an RL-driven framework that instead treats reusable skills themselves as the direct objects of evolution. Given hard-negative failures, EvoSkill-S1 selects parent skills and transforms them into candidate child skills through five genetic-inspired operators: revision, cross-revision, mutation, dual-reflection, and crossover. Downstream task rewards train the skill evolver to generate more effective child skills, while evaluator-guided write-back selectively retains beneficial updates. Complementary Skill Bank Fusion consolidates operator-specific banks to exploit their distinct recovery patterns, while a Procedural Skill Interface supports transfer to unseen problems. Across all reported math and code benchmarks, EvoSkill-S1 improves the corresponding initialized skill banks under both Qwen and GPT-family solvers, with relative gains of up to 12.39%. Compared with the strongest external baseline without direct skill evolution, the largest relative improvement reaches 20.39%.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.22610 [cs.AI] |
| (or arXiv:2611.22610v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.22610 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yujun Wang [view email]
[v1]
Tue, 17 Nov 2026 11:25:25 UTC (1,169 KB)
来源:arXiv cs.AI · arxiv.org