Accounting for the Sequential Nature of States to Learn Features for Reinforcement Learning - 42Papers