Saurabh Vyas is a postdoctoral research scientist in the Mortimer B. Zuckerman Mind Brain Behavior Institute at Columbia University working with Mark Churchland. His research on motor control broadly leverages a computation through neural population dynamics framework. Vyas’ Ph.D. research at Stanford University was advised by Krishna Shenoy in the Neural Prosthetic Systems Lab.

Vyas used a combination of brain-computer interfaces, large-scale neural population electrophysiology, and dynamical systems theory to explore how practicing a motor skill in the mind can speed up subsequent attempts of learning how to actually perform that same skill. He discovered that “covert” motor learning – that which is performed directly using neural activity and in the absence of physical movement – can directly facilitate overt learning. By recording and manipulating neural activity directly in premotor and primary motor cortex of nonhuman primates, Vyas established a causal relationship between preparatory activity – neural activity before the onset of movement – and motor learning. That is, covert learning systematically drives changes to preparatory neural activity, and critically, these changes persist across changes in context, thus driving transfer of learning. His experiments also made critical strides in expanding theoretical models of error-driven learning; these models were previously blind to the surprising influence of motor preparation on trial-by-trial motor learning. Together, his work suggests a potential avenue for neurorehabilitation through brain-computer interfaces, especially in cases where overt movement is not possible.