Effects of Spectral Normalization in Multi-agent Reinforcement Learning
This addresses stability issues in multi-agent on-policy learning for researchers and practitioners, but it is incremental as it applies an existing regularization technique to a specific problem.
The paper tackled the challenge of learning a reliable critic in multi-agent reinforcement learning with sparse rewards by applying spectral normalization regularization, resulting in faster and more robust learning in complex domains like SMAC and RWARE.
A reliable critic is central to on-policy actor-critic learning. But it becomes challenging to learn a reliable critic in a multi-agent sparse reward scenario due to two factors: 1) The joint action space grows exponentially with the number of agents 2) This, combined with the reward sparseness and environment noise, leads to large sample requirements for accurate learning. We show that regularising the critic with spectral normalization (SN) enables it to learn more robustly, even in multi-agent on-policy sparse reward scenarios. Our experiments show that the regularised critic is quickly able to learn from the sparse rewarding experience in the complex SMAC and RWARE domains. These findings highlight the importance of regularisation in the critic for stable learning.