领域感知智能体技能检索

中文摘要 AI 辅助

随着终身学习智能体积累不断增长的技能库，检索正确技能成为日益重要的瓶颈。当前大多数技能检索方法通过将名称、描述和主体等字段拼接，把每个技能视为一个扁平文档。然而，技能本质上是结构化的多字段对象，每个字段提供了关于技能使用时机和方式的不同信息。本研究探讨保留这种结构是否能改进技能检索。我们将每个技能表示为其独立组件，分别计算每个字段的稀疏和密集相似度，从而得到技能库的自然张量化、领域感知表示。随后我们将这些字段级分数结合，要么使用均匀权重，要么使用小型学习型 MLP（多层感知机）。在 SkillRet 和 SRA-Bench 两个不同的技能检索基准上，我们发现分离字段可改进混合检索，而对字段级分数进行学习能得到最强且最一致的结果。我们的领域感知 MLP 在 SkillRet 上达到 77.95 的 Recall@10（Top-10 召回率），在 SRA-Bench 上达到 83.78 的 Recall@10，优于对应的拼接学习基线。我们还发现，随着技能库变大，该优势会增长，表明领域感知技能检索在检索最困难的场景中尤其有用。我们的结果表明，技能表示本身很重要，仅保留技能文件中已有的结构就能大幅提升检索效果。

英文摘要

As lifelong learning agents accumulate lifelong growing skill banks, retrieving the correct skill becomes an increasingly important bottleneck. Most current skill retrieval methods treat each skill as one flat document by concatenating fields such as the name, description, and body. However, skills are naturally structured, multi-field objects, where each field provides different information about when and how the skill should be used. In this work, we study whether preserving this structure improves skill retrieval. We represent each skill as its separate components, and compute sparse and dense similarities for each field independently, exposing a naturally tensorized, field-aware representation of the skill bank. We then combine these field-level scores either with uniform weights or with a small learned MLP. Across two different skill retrieval benchmarks, SkillRet and SRA-Bench, we find that keeping fields separate improves hybrid retrieval, and learning over the field-level scores gives the strongest and most consistent results. Our field-aware MLP reaches $77.95$ Recall@10 on SkillRet and $83.78$ Recall@10 on SRA-Bench, outperforming the corresponding concatenated learned baselines. We also find that the advantage grows as the skill bank becomes larger, suggesting that field-aware skill retrieval becomes especially useful in the setting where retrieval is most difficult. Our results show that skill representation itself matters, and that simply preserving the structure already present in skill files can substantially improve retrieval.

发表机构

• University of California, Riverside（加州大学河滨分校）
• Nokia（诺基亚公司）

机构由 AI 辅助整理，请以论文原文为准。