COBRA-Skills: Contextual Bandit-Guided Evolution for Agent Skill Optimization· COBRA-Skills: 代理技能优化新框架
Large language model (LLM) agents can benefit from reusable skills distilled from prior task experience, yet existing skill optimization methods often rely on costly execution-based evaluation and substantial task data. We introduce \textbf{COBRA-Skills}, an efficient framework that formulates skill optimization as budgeted sequential optimization over a dynamically evolving candidate space. COBRA-Skills couples contextual-bandit-guided prioritization with evidence-grounded skill evolution, selectively allocating evaluations to promising or informative candidates while continually refining the skill population from execution feedback. Across six heterogeneous agent benchmarks and three target models, COBRA-Skills consistently achieves the strongest average performance among compared method
提出COBRA-Skills框架，高效优化代理技能。
- 核心方法
- 结合上下文多臂 bandit 指导的优先级分配与基于证据的技能进化，动态选择有潜力或信息丰富的候选技能进行评估。
- 适合谁读
- 研究者、工程师
- 要解决的问题
- 现有的代理技能优化方法依赖于高成本的执行评估和大量任务数据。
- 关键实验
- 关键实验包括在六个异构代理基准（如导航、问答等任务）和三个目标模型上进行测试，验证了框架的有效性。
- 主要贡献
- 在六个异构代理基准和三个目标模型上，COBRA-Skills 实现了所有比较方法中最强的平均性能。
- 意义与局限
- COBRA-Skills 提供了一种高效的代理技能优化方法，减少了对大量数据和计算资源的依赖，但可能在某些特定任务上的效果仍有待进一步验证。