Dopamine neurons encode a multidimensional probabilistic map of future reward
M. Sousa, P. Bujalski, B. F. Cruz, K. Louie, D. McNamee, J. J. Paton
This study introduces time-magnitude RL (TMRL), a new reinforcement learning framework, and shows that dopamine neurons encode a probabilistic map of future rewards in mice.
bioRxiv · Nov 12, 2023 · PreprintAnimal Behavior & Cognition