Multi-Agent Imitation by Learning and Sampling from Factorized Soft Q-Function | Read Paper on Bytez