Causal Multi-Agent Reinforcement Learning: Review and Open Problems
Machine Learning
2021-12-02 v2 Artificial Intelligence
Machine Learning
Abstract
This paper serves to introduce the reader to the field of multi-agent reinforcement learning (MARL) and its intersection with methods from the study of causality. We highlight key challenges in MARL and discuss these in the context of how causal methods may assist in tackling them. We promote moving toward a 'causality first' perspective on MARL. Specifically, we argue that causality can offer improved safety, interpretability, and robustness, while also providing strong theoretical guarantees for emergent behaviour. We discuss potential solutions for common challenges, and use this context to motivate future research directions.
Cite
@article{arxiv.2111.06721,
title = {Causal Multi-Agent Reinforcement Learning: Review and Open Problems},
author = {St John Grimbly and Jonathan Shock and Arnu Pretorius},
journal= {arXiv preprint arXiv:2111.06721},
year = {2021}
}
Comments
Accepted at Cooperative AI Workshop, NeurIPS 2021