SkillFM：通过潜在流匹配为LLM智能体生成技能
SkillFM: Generating Skills for LLM Agents via Latent Flow Matching
- Nanyang Technological University（南洋理工大学）
- University of Edinburgh（爱丁堡大学）
- Meta
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
提出SkillFM生成式框架，通过潜在流匹配直接合成任务条件文本技能，避免检索和强化学习，在ALFWorld和Search-QA上取得最佳性能。
AI中文摘要：
文本技能为大型语言模型智能体提供了可复用的指导，但现有方法通常依赖人工策划的技能库或使用间接且延迟反馈的强化学习。我们提出了SkillFM（技能流匹配），一个生成式框架，能够直接合成任务条件下的文本技能，无需测试时技能检索。我们的框架结合了一个用于在连续潜在空间中编码和重建文本技能的解码器，以及一个使用改进的MeanFlow训练的条件流模型。在推理时，学习到的速度场支持单步潜在采样，基于LLM的解码器将采样表示转换为冻结下游智能体的文本指导。我们在具身任务、问答和网页购物上评估了该框架。在ALFWorld和Search-QA上，我们的方法在比较的基于向量的技能方法中取得了最佳整体性能。我们的分析进一步表明，潜在技能生成是检索式技能增强的有效替代方案。我们的代码和训练技能库可在该https URL获取。
英文摘要：
Textual skills provide reusable guidance for large language model agents, but existing approaches often rely on manually curated skill banks or reinforcement learning with indirect and delayed feedback. We introduce SkillFM (Skill Flow Matching), a generative framework that synthesizes task-conditioned textual skills directly without test-time skill retrieval. Our framework combines a codec for encoding and reconstructing textual skills in a continuous latent space with a conditional flow model trained using improved MeanFlow. At inference time, the learned velocity field enables single-step latent sampling, and an LLM-based decoder converts the sampled representation into textual guidance for a frozen downstream agent. We evaluate the framework on embodied tasks, question answering, and web shopping. On ALFWorld and Search-QA, our method achieves the best overall performance among the compared vector-based skill approaches. Our analyses further demonstrate that latent skill generation is an effective alternative to retrieval-based skill augmentation. Our code and training skill libraries are available at https://github.com/lulushang999/SkillFM.
