Lizmaylin Ramos

Abstract To learn motivated behaviors, animals must both anticipate rewards and reinforce actions that yield them. Although cerebellar activity encodes natural rewards such as water and food, it also coordinates physical movements such as eating and drinking. To disentangle reward from consummatory movements, we trained mice to push for delayed dopamine (DA) rewards delivered directly into the brain. Here we show that cerebellar input streams use both predictive and instructive codes for DA reward.

Abstract To learn effectively, animals must generalize across related contexts yet distinguish between them. Generalization relies on low-dimensional neural manifolds throughout the neocortex1,2, which accelerate learning by constraining neural activity to task-relevant axes3.Conversely, context separation is thought to depend on neural expansion layers that can project information into high-dimensional feature spaces4,5, most famously cerebellar granule cells (GrCs)6,7,8.