The brain exhibits a remarkable range of flexibility and inflexibility. A key insight from reinforcement learning (RL) models is that humans’ ability flexibly to plan new actions—and also our failures ...