Learning to Use Tools via Cooperative and Interactive Agents

TL;DR

This paper introduces ConAgents, a framework with three specialized, cooperative agents for tool use in language models, significantly improving task success rates over existing single-agent methods.

Contribution

The paper proposes a novel multi-agent framework with communication protocols and action distillation to enhance tool use and adaptability in language models.

Findings

Up to 14% higher success rate on three datasets

Effective coordination among specialized agents improves performance

Action distillation enhances generalization to open-source models

Abstract

Tool learning empowers large language models (LLMs) as agents to use external tools and extend their utility. Existing methods employ one single LLM-based agent to iteratively select and execute tools, thereafter incorporating execution results into the next action prediction. Despite their progress, these methods suffer from performance degradation when addressing practical tasks due to: (1) the pre-defined pipeline with restricted flexibility to calibrate incorrect actions, and (2) the struggle to adapt a general LLM-based agent to perform a variety of specialized actions. To mitigate these problems, we propose ConAgents, a Cooperative and interactive Agents framework, which coordinates three specialized agents for tool selection, tool execution, and action calibration separately. ConAgents introduces two communication protocols to enable the flexible cooperation of agents. To effectively generalize the ConAgents into open-source models, we also propose specialized action distillation, enhancing their ability to perform specialized actions in our framework. Our extensive experiments on three datasets show that the LLMs, when equipped with the ConAgents, outperform baselines with substantial improvement (i.e., up to 14% higher success rate).