Coordinated cerebellar climbing fiber activity signals learned sensorimotor predictions

The prevailing model of cerebellar learning states that climbing fibers (CFs) are both driven by, and serve to correct, erroneous motor output. However, this model is grounded largely in studies of behaviors that utilize hardwired neural pathways to link sensory input to motor output. To test whether this model applies to more flexible learning regimes that require arbitrary sensorimotor associations, we have developed a cerebellar-dependent motor learning paradigm compatible with both mesoscale and single dendrite resolution calcium imaging in mice. Here, we find that CFs are preferentially driven by and more time-locked to correctly executed movements and other task parameters that predict reward outcome, exhibiting widespread correlated activity within parasagittal processing zones that is governed by these predictions. Together, such CF activity patterns are well-suited to drive learning by providing predictive instructional input consistent with an unsigned reinforcement learning signal that does not rely exclusively on motor errors.