Towards Causal Reinforcement Learning - Part II