Dynamic organization of cerebellar climbing fiber response and synchrony in multiple functional modules reduces dimensions for reinforcement learning

Abstract

Daynamic functional organization by synchronization is theorized to be essential for dimension reduction of the cerebellar learning space. We analyzed a large amount of coordinate-localized, two-photon imaging data from cerebellar Crus II in mice undergoing "Go/No-go" reinforcement learning. Tensor component analysis revealed that a majority of climbing fiber inputs to Purkinje cells were reduced to only four functional components, corresponding to accurate timing control of motor initiation related to a Go cue, cognitive error-based learning, reward processing, and inhibition of erroneous behaviors after a No-go cue. Spatial distribution of these components coincided well with the boundaries of Aldolase-C/zebrin II expression in Purkinje cells, whereas several components are mixed in single neurons. Synchronization within individual components was bidirectionally regulated according to specific task contexts and learning stages. These findings suggest that the cerebellum, based on anatomical compartments, reduces dimensions by self-organization of components, a feature that may inspire new-generation AI designs.