Home

Awesome

Suspicion Agent: Playing Imperfect Information Games with Theory of Mind Aware GPT-4

The implementation of "Suspicion Agent: Playing Imperfect Information Games with Theory of Mind Aware GPT-4".

Project Demo | Game Data Replay

Method

figure figure

figure

Instruction

Install required packages with below command (python >= 3.8.5):

pip install -r requirements.txt

Set up the OpenAI key following OPENAI_KEY

Train and evaluate agents:

python main_vs_baseline.py --user --verbose_print  --rule_model [cfr/ nfsp / dqn / dmc]

Sample Output

figure figure

Star History

Star History Chart

Reference

@misc{guo2023suspicionagent,
      title={Suspicion-Agent: Playing Imperfect Information Games with Theory of Mind Aware GPT4}, 
      author={Jiaxian Guo and Bo Yang and Paul Yoo and Yuchen Lin and Yusuke Iwasawa and Yutaka Matsuo},
      year={2023},
      eprint={2309.17277},
      archivePrefix={arXiv},
      primaryClass={cs.AI}
}

Note: this code is based on the SkyAGI. Many thanks to the authors.