Toward a manifold encoding neural responses in the visual system

Thursday, July 20, 2023 14:00to15:00
MGH L7 140, CA

BRaIN Seminar Series

Brain Repair and Integrative Neuroscience Program The BRaIN Program of the RI-MUHC is pleased to announce our next presentation under the theme:Vision and Cognitive Neuroscience

Speaker: Dr. Steven Zucker

David & Lucile Packard Professor of Computer Science & Biomedical Engineering
Yale University, School of Engineering & Applied Science
Founding member of the Centre for Intelligent Machines (CIM)
How might one infer circuit properties from neurophysiological data? How do these circuits relate to artificial neural networks?
We address these challenges with a novel neural manifold. It is obtained using unsupervised machine learning algorithms and applied to the mouse visual system. Each point on our manifold is a neuron; nearby neurons respond similarly in time to similar parts of a stimulus ensemble. This ensemble includes drifting gratings and flows, i.e. patterns resembling what a mouse would "see" while running through fields. Our manifold differs from the standard practice in computational neuroscience, of embedding trials in neural coordinates. Importantly, for our manifolds topology matters: from spectral theory we infer that, if the circuit consists of separate components, the manifold is discontinuous (illustrated with retinal data). If there is significant overlap between circuits, the manifold is nearly-continuous (cortical data).  To approach real circuits, local neighborhoods on the manifold are identified with actual circuit components. For the retinal data we show these components correspond to distinct ganglion cell types by their mosaic-like receptive field organization, while for cortical data, neighborhoods organize neurons by type (excitatory/inhibitory) and anatomical layer. The manifold topology for deep CNN's will also be developed.

Joint research with Luciano Dyballa (Yale), Marija Rudzite (Duke), Michael Styrker (UCSF) and Greg Field (UCLA).
Back to top