Reinforcement Learning, Bit by Bit - 42Papers