Mixture Weighted Policy Cover: Exploration In Multi-Agent Reinforcement Learning