Tool Graph Retriever: Exploring Dependency Graph-based Tool Retrieval for Large Language Models

2025-08-07

AI Summary

In large language model (LLM)-based agents, the proliferation of tools causes context overload, and existing semantic-similarity-based tool retrieval methods neglect inter-tool dependencies, often omitting critical prerequisite tools. Method: This work introduces the novel concept of a “Tool Dependency Graph” (TDG), constructs the TDI300K dataset to train a dependency discriminator, and designs a Graph Convolutional Network (GCN) to explicitly model candidate tools as a graph and iteratively refine their representations—jointly encoding both semantic and structural dependency information. Contribution/Results: The proposed method enables online, efficient, dependency-aware tool retrieval. It significantly improves tool recall and task success rates across multiple benchmark datasets, achieving state-of-the-art (SOTA) performance. Empirical results demonstrate that explicit modeling of tool dependencies yields substantial and consistent gains in retrieval quality, validating its critical role in robust LLM-agent tool orchestration.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Multiagent Systems: Modeling other Agents

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

With the remarkable advancement of AI agents, the number of their equipped tools is increasing rapidly. However, integrating all tool information into the limited model context becomes impractical, highlighting the need for efficient tool retrieval methods. In this regard, dominant methods primarily rely on semantic similarities between tool descriptions and user queries to retrieve relevant tools. However, they often consider each tool independently, overlooking dependencies between tools, which may lead to the omission of prerequisite tools for successful task execution. To deal with this defect, in this paper, we propose Tool Graph Retriever (TGR), which exploits the dependencies among tools to learn better tool representations for retrieval. First, we construct a dataset termed TDI300K to train a discriminator for identifying tool dependencies. Then, we represent all candidate tools as a tool dependency graph and use graph convolution to integrate the dependencies into their representations. Finally, these updated tool representations are employed for online retrieval. Experimental results on several commonly used datasets show that our TGR can bring a performance improvement to existing dominant methods, achieving SOTA performance. Moreover, in-depth analyses also verify the importance of tool dependencies and the effectiveness of our TGR.

Problem

• Efficient retrieval of tools for large language models
• Overcoming omission of prerequisite tool dependencies
• Improving tool representations using dependency graphs

Innovation

• Constructs TDI300K dataset for tool dependencies
• Uses graph convolution for tool representation
• Integrates tool dependencies into retrieval process