Low Entropy Communication in Multi-Agent Reinforcement Learning | Read Paper on Bytez