Multi-agent reinforcement learning framework with permutation invariant and equivariant networks...