SKIMIX：面向动态 harness 工程的技能混合多智能体 harness 时间缩放方法
SKIMIX: Multi-Agent Harness-Time Scaling with Skill Mixture for Dynamic Harness Engineering
浏览论文内容
中文总结 AI 辅助
SKIMIX 是一种多智能体框架，结合技能检索、抗稀释路由等技术，在六个推理基准上提升开放式数学推理能力，为可扩展智能体设计提供了任务特性相关的实用指导。
中文摘要 AI 辅助
AI 智能体越来越依赖大型技能库，但技能的选择、组合与维护仍存在困难。我们提出 SKIMIX，这是一种多智能体框架，其中具有不同技能组合的智能体通过迭代优化进行协作。SKIMIX 结合了基于嵌入的技能检索、子模抗稀释路由以及自适应技能演化。在六个推理基准测试中，多智能体协作显著提升了开放式数学推理能力，但在选择题任务上仅提供有限增益或出现负增益。智能体数量的缩放呈非单调变化，且大部分改进出现在第一轮优化期间。这些结果表明，任务特性决定了技能级集成是否有效，并为可扩展的智能体设计提供了实用指导。
英文摘要
AI agents increasingly rely on large skill libraries, but selecting, combining, and maintaining skills remains difficult. We propose SKIMIX, a multi-agent framework in which agents with different skill portfolios collaborate through iterative refinement. SKIMIX combines embedding-based skill retrieval, submodular anti-dilution routing, and adaptive skill evolution. Across six reasoning benchmarks, multi-agent collaboration substantially improves open-ended mathematical reasoning but offers limited or negative gains on multiple-choice tasks. Agent-count scaling is non-monotonic, and most improvements arise during the first refinement round. These results show that task characteristics determine whether skill-level ensembles help and provide practical guidance for scalable agent design.
