Causal Multi-Agent Reinforcement Learning: Review and Open Problems
It addresses challenges in MARL for researchers and practitioners by proposing causal methods as a solution, but it is incremental as it reviews and motivates rather than presents new results.
This paper reviews the intersection of multi-agent reinforcement learning (MARL) and causality, arguing that a 'causality first' approach can improve safety, interpretability, and robustness in MARL systems.
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.