# Large Language Model based Multi-Agents: A Survey of Progress and Challenges

**Authors:** Taicheng Guo, Xiuying Chen, Yaqi Wang, Ruidi Chang, Shichao Pei, Nitesh V. Chawla, Olaf Wiest, Xiangliang Zhang  
**arXiv:** [2402.01680](https://arxiv.org/abs/2402.01680)  
**Date:** 2024-01-21  
**Summary page:** https://www.summarizepaper.com/en/paper/2402.01680v1/

## Key points

- Large Language Models (LLMs) are valuable autonomous agents for automating tasks due to their advanced planning and reasoning capabilities.
- Utilizing LLMs as decision-making agents within multi-agent systems has led to significant progress in complex problem-solving and world simulation.
- The survey aims to provide a comprehensive overview of multi-agent systems based on LLMs by discussing essential aspects and addressing associated challenges.
- Key questions explored include domains and environments simulated by LLM-based multi-agents, agent profiling and communication, and mechanisms enhancing agents' capacities.
- The authors maintain an open-source GitHub repository dedicated to outlining the latest studies on LLM-based multi-agent systems, inviting contributions from the community.

## Summary

In their work titled "Large Language Model based Multi-Agents: A Survey of Progress and Challenges," authors Taicheng Guo, Xiuying Chen, Yaqi Wang, Ruidi Chang, Shichao Pei, Nitesh V. Chawla, Olaf Wiest, and Xiangliang Zhang delve into the realm of Large Language Models (LLMs) and their application in multi-agent systems. LLMs have demonstrated exceptional success in various tasks due to their advanced planning and reasoning capabilities, making them valuable autonomous agents for automating tasks. The authors highlight the recent advancements in utilizing LLMs as decision-making agents within multi-agent systems, leading to significant progress in complex problem-solving and world simulation. Their survey aims to provide a comprehensive overview of this evolving field by discussing essential aspects of multi-agent systems based on LLMs and addressing the associated challenges. Readers are encouraged to explore key questions such as the domains and environments simulated by LLM-based multi-agents, how these agents are profiled and communicate, and the mechanisms contributing to enhancing agents' capacities. Additionally, the authors summarize commonly used datasets or benchmarks for those interested in further exploration. To facilitate ongoing research in this area,<Organization>the authors maintain an open-source GitHub repository dedicated to outlining the latest studies on LLM-based multi-agent systems</Organization>. They invite contributions from the community to enrich this dynamic field of study.