Tadashi Yamazaki

This paper develops a cerebellar spiking network as a reinforcement learning model in continuous time and space, demonstrating its ability to solve RL tasks and challenge the classical view of cerebellar learning.