Consensus Learning for Cooperative Multi-Agent Reinforcement Learning

Almost all multi-agent reinforcement learning algorithms without communication follow the principle of centralized training with decentralized execution. During training, agents can be guided by same signals, such as global state. However, lack shared signal and choose actions given local observations during Inspired viewpoint invariance contrastive learning, we propose consensus for cooperative in this study. Although based on observations, different infer discrete spaces communication. We feed inferred one-hot to network an explicit input a way, thereby fostering their spirit. With minor model modifications, our suggested framework extended variety algorithms. Moreover, carry out these variants some fully tasks get convincing results.