MetaAgent: Automatically Constructing Multi-Agent Systems Based on Finite State Machines

TL;DR

MetaAgent is a finite state machine framework that automatically constructs and optimizes multi-agent systems from task descriptions, outperforming existing auto-designed methods and matching human-designed systems in performance.

Contribution

The paper introduces MetaAgent, a novel finite state machine-based framework for automatic multi-agent system design from task descriptions, addressing limitations of prior automated methods.

Findings

MetaAgent outperforms other auto-designed methods.

MetaAgent achieves comparable performance to human-designed systems.

Effective in both text-based and practical tasks.

Abstract

Large Language Models (LLMs) have demonstrated the ability to solve a wide range of practical tasks within multi-agent systems. However, existing human-designed multi-agent frameworks are typically limited to a small set of pre-defined scenarios, while current automated design methods suffer from several limitations, such as the lack of tool integration, dependence on external training data, and rigid communication structures. In this paper, we propose MetaAgent, a finite state machine based framework that can automatically generate a multi-agent system. Given a task description, MetaAgent will design a multi-agent system and polish it through an optimization algorithm. When the multi-agent system is deployed, the finite state machine will control the agent's actions and the state transitions. To evaluate our framework, we conduct experiments on both text-based tasks and practical tasks. The results indicate that the generated multi-agent system surpasses other auto-designed methods and can achieve a comparable performance with the human-designed multi-agent system, which is optimized for those specific tasks.