SkillLift：从稀疏预言中学习密集评分标准以实现高效技能进化
SkillLift: Learning Dense Rubrics from Sparse Oracles for Efficient Skill Evolution
- Fudan University（复旦大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
SkillLift提出通过排序监督和双层优化，用稀疏预言学习密集评分标准，解耦技能搜索与预言成本，实现高效技能进化，以更少令牌成本超越现有方法。
AI中文摘要：
基于LLM的智能体日益依赖持久技能（即可复用的程序化提示）来在不更新权重的情况下进行适应。现有的技能自我进化方法直接根据执行反馈修改技能文本，但每次预言评估都需要完整的智能体回滚，这造成了监督瓶颈，将搜索限制在失败修补更新上。我们的关键见解是，排序比绝对结果回归是更平滑的监督目标：识别哪个技能更好所需的预言评估次数少于预测精确分数。基于这一见解，我们提出了SkillLift，它通过将技能搜索与预言成本解耦，学习一个与预言对齐的评分标准作为结构化评估空间。我们将其形式化为通过交替优化求解的双层优化问题：内循环使用冻结的评分标准作为廉价替代品，在零预言成本下指导技能修订；外循环调用少量预言回滚，通过秩相关重新对齐评分标准，从而分摊预言成本并稳定文本空间更新。在复杂智能体任务基准上的实验表明，我们的方法以比前沿进化方法少40%至70%的令牌成本超越了现有的自动技能方法。代码可在该https URL获取。
英文摘要：
LLM-based agents increasingly rely on persistent skills, i.e., reusable procedural prompts, to adapt without weight updates. Existing skill self-evolution methods directly revise skill text based on execution feedback, but each oracle evaluation requires a full agent rollout, creating a supervision bottleneck that confines search to failure-patching updates. Our key insight is that ranking is a smoother supervision target than absolute outcome regression: identifying which skill is better requires fewer oracle evaluations than predicting exact scores. Building on this insight, we propose SkillLift, which decouples skill search from oracle cost by learning an oracle-aligned rubric as a structured evaluation space. We formalize this as a bilevel optimization problem solved via alternating optimization: an inner loop uses the frozen rubric as a cheap surrogate to guide skill revision at no oracle cost, while an outer loop invokes a small number of oracle rollouts to re-align the rubric via rank correlation, amortizing oracle cost and stabilizing text-space updates. Experiments on complex agent task benchmarks show that our method outperforms existing auto-skill methods with 40--70\% less token cost compared to frontier evolving methods. Codes are available at https://github.com/WalteR-MittY-pro/SkillLift.
