Theory of Mind for Multi-Agent Collaboration via Large Language Models

June 11, 2024 ยท View on GitHub

This repository contains reference implementation for multi-LLM ToM paper (accepted to EMNLP 2023), Theory of Mind for Multi-Agent Collaboration via Large Language Models, available at https://aclanthology.org/2023.emnlp-main.13/

Cite

If you use this code in your work, please cite the following:

@inproceedings{li2023theory,
  title={Theory of Mind for Multi-Agent Collaboration via Large Language Models},
  author={Li, Huao and Chong, Yu and Stepputtis, Simon and Campbell, Joseph P and Hughes, Dana and Lewis, Charles and Sycara, Katia},
  booktitle={Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing},
  pages={180--192},
  year={2023}
}

Installation

First, install the USAR environment gym-dragon.

# downgrade setuptools for gym=0.21
pip install setuptools==65.5.0 "wheel<0.40.0"
cd gym-dragon
pip install -e .

Next, we need to install dependencies for LLM agents including openai API package. For doing that run:

pip install -r requirements.txt

Finally, we need to set up the openai API key in your system environment variables as instructed in this guideline.

echo "export OPENAI_API_KEY='yourkey'" >> ~/.bash_profile
source ~/.bash_profile
echo $OPENAI_API_KEY

Running

Once everything is installed, we can run the using these example commands

Urban Search and Rescue (i.e. gym_dragon)

  • GPT-4-turbo on mini_dragon with 5 nodes
python dragonExp.py --model gpt-4-turbo-preview --exp_name gpt-4 --allow_comm --belief --seed 0
  • GPT-4-turbo w/o communication on mini_dragon with 5 nodes
python dragonExp.py --model gpt-4-turbo-preview --exp_name gpt-4 --belief --seed 0
  • GPT-4-turbo on mini_dragon with 5 nodes and ToM inference tasks
python dragonExp.py --model gpt-4-turbo-preview --exp_name gpt-4 --allow_comm --ToM --belief --seed 0

Contributors

License

Code is available under MIT license.