How the Vestibulocerebellum Builds an Internal Model of Self-motion

The vestibulocerebellum transforms raw vestibular signals into central estimates of head motion. One well-studied transformation is the extraction of head tilt and translation from vestibular signals, which requires integrating signals from two different vestibular sensors: the semicircular canals and the otoliths. Here we review how tilt and translation are encoded in the vestibulocerebellum. We focus on computational aspects: we present the theoretical framework of tilt-translation discrimination, as well as the mathematical methods used to model how cerebellar neurons encode tilt and translation. We also present the experimental and statistical frameworks that allow identifying which variables are encoded by individual cerebellar neurons and probing into the underlying multisensory integration operations.