arXiv cs.CL· Anna Mazhar, Seiji Maekawa, Nikita Bhutani·· 12 小时前AI 评分38
BRIA：在大型工具库中分支探索、选择性采纳的执行引导工具使用智能体
Explore Separately, Adopt Selectively: Execution-Guided Tool Use over Large Tool Libraries
AI 导读
研究者提出智能体 BRIA（BRanch-Inspect-Adopt），将候选工具的调查隔离在独立分支中执行，任务智能体只把有用的执行结果采纳进主轨迹。
正文 · 原文
Abstract:Large language model agents must discover relevant tools from large libraries, e.g., thousands of tools, as new needs arise during task execution. When several candidates could satisfy the same need, execution may be required to identify the relevant one. Investigating candidates in a shared trajectory can leave alternatives unexplored or allow failed attempts to influence subsequent reasoning. To address this, we present BRIA (BRanch-Inspect-Adopt), a simple yet effective agent that temporarily isolates candidate investigations in separate branches. Each branch retrieves and executes tools to investigate its candidate; a task agent inspects the resulting execution evidence and adopts only useful results into its continuing trajectory. BRIA retrieves additional tools as new capability needs arise during task execution. We evaluate BRIA on PlanBench-XL and MCP-Atlas using a shared pool of 1897 tools, three language models, and static and adaptive retrieval baselines. Across all settings, BRIA matches or exceeds the strongest retrieval baseline. On MCP-Atlas, it achieves an average claim score of 89.77, compared with 75.40 for MCP-Zero, a gain of 14.37 percentage points. BRIA also lies on the Pareto frontier for task performance and input-token cost.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.22688 [cs.CL] |
| (or arXiv:2610.22688v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.22688 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Anna Mazhar [view email]
[v1]
Thu, 15 Oct 2026 21:56:10 UTC (557 KB)
来源：arXiv cs.CL · arxiv.org