Neocortex-cerebellum learning algorithms

5:00pm - 6:00pm
Querido Hall & online
Speakers Website
Dr. Mark Wagner

Abstract Cortex and cerebellum have jointly expanded over mammalian brain evolution, are linked by universal reciprocal connections, and together contain 99% of all neurons in humans. My lab’s central goal is to identify a general-purpose cortico-cerebellar learning algorithm which can support diverse functional domains via common circuit motifs. Our approaches include chronic multi-site two-photon imaging and optogenetics, electrophysiology, and computational modeling. In one line of inquiry, we aim to understand cerebellar contributions to nonmotor behavior, including reward-driven learning, to identify processes common to both motor and nonmotor cortico-cerebellar function. Second, we are developing new ways to probe the dialogues across cortico-cerebellar network elements, to understand how they cooperate to achieve learning goals. Third, we are using our physiology data to constrain new computational models of cortico-cerebellar circuit function, with the long-term goal of explaining how this circuit architecture supports a variety of both motor and nonmotor learning processes.


Xiaolu Wang, Vincenzo Romano,