MetaToolAgent: Towards Generalizable Tool Usage in LLMs through Meta-Learning
This addresses the challenge of dynamic tool coordination for LLMs in practical deployments, though it is incremental as it builds on existing tool learning methods.
The paper tackles the problem of limited generalization in tool selection for large language models (LLMs) by introducing MetaToolAgent, a meta-learning approach that significantly outperforms baselines on unseen tools.
Tool learning is increasingly important for large language models (LLMs) to effectively coordinate and utilize a diverse set of tools in order to solve complex real-world tasks. By selecting and integrating appropriate tools, LLMs extend their capabilities beyond pure language understanding to perform specialized functions. However, existing methods for tool selection often focus on limited tool sets and struggle to generalize to novel tools encountered in practical deployments. To address these challenges, we introduce a comprehensive dataset spanning 7 domains, containing 155 tools and 9,377 question-answer pairs, which simulates realistic integration scenarios. Additionally, we propose MetaToolAgent (MTA), a meta-learning approach designed to improve cross-tool generalization. Experimental results show that MTA significantly outperforms baseline methods on unseen tools, demonstrating its promise for building flexible and scalable systems that require dynamic tool coordination.