Apple 研究：智能体先规划还是先检索工具更有效
Plan or Retrieve First? A Scale Study of Tool Discovery Architectures
Apple 把智能体找工具的两种思路拆开对比了，讲了 RtP 和 PtR 各自怎么运作、规模化后差在哪，做 Agent 开发的话能直接参考。
Apple 的研究对比了工具发现的两种架构：Retrieve-then-Plan（RtP）先用用户查询做近邻检索，Plan-then-Retrieve（PtR）则让 LLM 先生成 Abstract Tool Stubs 作为检索查询。由于上下文窗口有限，无法把所有工具定义塞进单一提示词，架构选择直接影响智能体在大型工具注册表中的表现。该研究在多个基准上评估了两种范式的规模化表现差异。
Plan or Retrieve First? A Scale Study of Tool Discovery Architectures
Tool-augmented LLM agents discover relevant APIs from large registries before planning execution, yet finite context windows prohibit injecting all tool definitions into a single prompt. The dominant approach, which we term Retrieve-then-Plan (RtP), embeds the user query and retrieves similar tools via approximate nearest-neighbor search. We study an alternative Plan-then-Retrieve (PtR) paradigm in which the LLM first generates Abstract Tool Stubs (hypothesized function signatures describing needed capabilities) which then serve as retrieval queries against the registry. Evaluating on…
