Grounded in reinforcement learning. General idea is hippocampal replay is valuable, and a rational agent should arbitrate between work and replay in order to ma

Grounded in reinforcement learning. General idea is hippocampal replay is valuable, and a rational agent should arbitrate between work and replay in order to maximize future reward