Alex Huk (UCLA)

Title: Population-level insights into primate vision and cognition

Abstract: Many foundational insights into the neural basis of perception and cognition have come from single-neuron recordings in awake primates, resulting in classical computational models in domains including vision and visually-guided working memory. However, the difficulty of deploying large-scale recording techniques (as well as circuit- and cell-type focused tools) has slowed attempts to take such phenomenological models and refine them based on detailed measurements of brain function (which, I would argue, is the point of doing neurophysiology). Here, we show that large-scale recordings in the brains of awake, behaving marmosets allow for insights into population-level representations and computations that both refine existing models and call for re-evaluations of some of their assumptions. I will describe ongoing work in my lab exploring: (1) how behavioral state may affect visual processing (in early visual areas V1 and V2); (2) how vision and action interact during movements (in mid-level visual area MT); and (3) how sensory events are transformed into temporally-persistent information for cognition and motor planning (in sensorimotor areas LIP and FEF). These studies reflect a first step in turning classic computational models into real biological theories, and also begin building a bridge to insights from deploying powerful techniques in other model systems, such as mice.

Host: Spencer Smith