arXiv cs.AI· Kounianhua Du, Jianxing Liu, Ruoshui Shan, Yucheng Zhang, Ke Li, Xing Sun, Weinan Zhang·· 12 小时前AI 评分31
MultiEvoSkill：面向可泛化技能产物的结构感知多智能体技能进化框架
MultiEvoSkill: Structure-Aware Multi-Agent Skill Evolution for Generalizable Skill Artifacts
AI 导读
MultiEvoSkill 是一个技能优化框架，通过技能编译器将技能转为带类型图结构、异构智能体系统基于多组经验提出修订建议、结构感知辩论与进化机制在依赖和重要性约束下整合候选编辑，解决单一模型视角技能难以跨模型泛化、顺序描述无法表达并行分支与依赖的问题。多维门控确保每次进化的信息增益与拓扑正确性，系统收敛时得到最优技能。对比 SOTA 方法的实验与消融验证了其有效性与泛化能力。
正文
Abstract:Skill documents are becoming the default adaptation surface for frozen large language model (LLM) agents, but current skill generators face two key limitations. First, humans and different LLMs differ in how they organize procedural knowledge and approach task solving. Skills authored from a human or single-model perspective may therefore fail to align with the reasoning behaviors of other backbones, limiting cross-model generalization. Second, sequential skill descriptions inadequately represent procedures with parallel branches and explicit dependencies, while existing skill optimizers primarily operate on prose and lack explicit support for graph-structured skill evolution. We introduce MultiEvoSkill, a skill-optimization framework comprising three core components: 1) Skill compiler casts the skill as a typed graph with tailored operators, producing a compact, informative representation with explicit procedural topology; 2) Heterogeneous agent system proposes revision advices using collective intelligence, where minimal and informative issues are identified based on multiple sets of experiences. 3) Structure-aware debate and evolution mechanism organizes candidate edits into a heterogeneous graph, resolves them under dependency and importance constraints, and distills the collective proposal into the artifact. Multi-dimensional gating ensures the information gain and topological correctness of each evolution. Optimal skills are then obtained when the system converges at a consistent state. Various experiments and ablations against SOTA methods validate the effectiveness and generalization of MultiEvoSkill.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.34260 [cs.AI] |
| (or arXiv:2610.34260v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.34260 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jianxing Liu [view email]
[v1]
Sun, 25 Oct 2026 12:43:01 UTC (3,408 KB)
来源：arXiv cs.AI · arxiv.org