This repository accompanies our paper "Mediated Multi-Agent Reinforcement Learning" published at AAMAS, 2023.
This is the code we used for our experiments in one of the earlier revisions of the paper. It is not good. We tried to make a more general codebase but it ended up taking too much time. We're releasing this because it's better than nothing.
The minimial working example is in a single notebook in the pgg folder, but note that it has some ad-hoc solutions for the sake of simplicity. If you want to apply mediators to your environment, the most general version of the code is in pgg-iter.
If you use our code, please cite us:
@inproceedings{ivanov2023mediated,
title={Mediated Multi-Agent Reinforcement Learning},
author={Ivanov, Dmitry and Zisman, Ilya and Chernyshev, Kirill},
booktitle={Proceedings of the 2023 International Conference on Autonomous Agents and Multiagent Systems},
pages={49--57},
year={2023}}