Multi-Field Tool Retrieval

TL;DR

This paper introduces Multi-Field Tool Retrieval, a framework that improves the accuracy and robustness of retrieving external tools for LLMs by modeling multiple aspects of tool utility, addressing key challenges in documentation and query matching.

Contribution

It proposes a novel multi-field modeling approach for tool retrieval, significantly enhancing performance and generalizability over existing ad-hoc methods.

Findings

Achieves state-of-the-art results on five datasets and a mixed benchmark.

Demonstrates superior robustness and generalizability.

Effectively aligns user intent with tool representations.

Abstract

Integrating external tools enables Large Language Models (LLMs) to interact with real-world environments and solve complex tasks. Given the growing scale of available tools, effective tool retrieval is essential to mitigate constraints of LLMs' context windows and ensure computational efficiency. Existing approaches typically treat tool retrieval as a traditional ad-hoc retrieval task, matching user queries against the entire raw tool documentation. In this paper, we identify three fundamental challenges that limit the effectiveness of this paradigm: (i) the incompleteness and structural inconsistency of tool documentation; (ii) the significant semantic and granular mismatch between user queries and technical tool documents; and, most importantly, (iii) the multi-aspect nature of tool utility, that involves distinct dimensions, such as functionality, input constraints, and output formats, varying in format and importance. To address these challenges, we introduce Multi-Field Tool Retrieval, a framework designed to align user intent with tool representations through fine-grained, multi-field modeling. Experimental results show that our framework achieves SOTA performance on five datasets and a mixed benchmark, exhibiting superior generalizability and robustness.