The successor representation was introduced into reinforcement learning by Dayan ( 1993 ) as a means of facilitating generalization between states with similar successors. Although reinforcement learning in general has been used extensively as a model of psychological and neural processes, the ps…