Research Overview

The central aim of our research is to understand the cerebellum’s role in behaviour.

The cerebellum is a unique region because of its evolutionarily-conserved, structured architecture and key role in motor learning. Inspired by its so-called ‘crystalline’ circuitry, classic theories have argued that the cerebellum performs sensorimotor prediction in order to coordinate, adapt, and fine-tune movements during learning.

However, over the past several years, accumulating evidence in humans and mice has hinted that the cerebellum plays a subtle, yet important, role in cognitive processing as well. How can the conserved cerebellar circuitry, which for decades has been studied under the lens of motor control, support cognitive behaviours? We are working to close this gap by studying cerebellar function at the circuit and systems levels.

Towards this end, we use a variety of quantitative methods, including (but not limited to): dimensionality reduction, RNNs, reinforcement learning, stochastic modelling, and dynamical systems. We do both pure modelling as well as data analysis. For the latter, we work closely with our experimental collaborators to analyze large-scale neural recordings and behavioral data.