Embodied LLM Agents Learn to Cooperate in Organized Teams

Source codes for the paper:

Guo, Xudong, Kaixuan Huang, Jiale Liu, Wenhui Fan, Natalia Vélez, Qingyun Wu, Huazheng Wang, Thomas L. Griffiths, and Mengdi Wang: Embodied LLM Agents Learn to Cooperate in Organized Teams

Paper

Website

In this repo, we implement embodied and organized multi-LLM-agent teams in the environment VirtualHome to cooperate on household tasks. The agents can make decisions and communicate freely with each other based on LLMs. The codes support multi-agent communication (>3 agents) such as broadcasting, keeping silent, and selecting receivers to send distinct messages.

Thanks to the flexible communication protocol, the organizational structures can be updated accordingly and autonomously by editing the organization prompts. The agents show emergent cooperative behaviors such as information sharing, leadership & assistance, and requests for guidance.

To improve the organization prompts, we propose the Criticize-Reflect architecture based on LLMs to analyze the trajectories and reflect on the feedback to generate novel prompts.

Installation

Clone the VirtualHome API repository:

git clone --branch wah https://github.com/xavierpuigf/virtualhome.git

Download the Simulator (Linux x86-64 version), and unzip it.

The files should be organized as follows:

|--cwah/
|--virtualhome/
|--executable/

Experiments

Set up your API keys or local LLM models in /envs/cwah/scripts /llm_configs.json

"gpt-4" : 
    {
        "api_key": "API_KEY",
        "model": "gpt-4"
    }

Add your organization prompts in /envs/cwah/testing_agents /organization_instructions.csv

For example,

Agent 1 is the leader to coordinate the task.

To run the experiment with a fixed organization prompt:

bash envs/cwah/scripts/fixed_organization.sh

To improve the organization prompt by Criticize-Reflect:

bash envs/cwah/scripts/criticize_reflect.sh

Check the logs and the visualization of communication in the folder envs/cwah/log

Other core files:

Interface between agents and envs: envs/cwah/algos/arena_mp2.py

Communication module based on AutoGen: envs/cwah/algos/comms.py

LLM agent: envs/cwah/agents/LLM_agent.py

LLM and prompts: envs/cwah/LLM/LLM.py

Citation

@article{guo2024embodied,
  title={Embodied LLM Agents Learn to Cooperate in Organized Teams},
  author={Guo, Xudong and Huang, Kaixuan and Liu, Jiale and Fan, Wenhui and V{\'e}lez, Natalia and Wu, Qingyun and Wang, Huazheng and Griffiths, Thomas L and Wang, Mengdi},
  journal={arXiv preprint arXiv:2403.12482},
  year={2024}
}

Acknowledgements

Our work is based on the repos: Co-LLM-Agents and AutoGen

rui-qian/Organized-LLM-Agents

Embodied LLM Agents Learn to Cooperate in Organized Teams

Installation

Experiments

Citation

Acknowledgements