CCN Seminar, Evan Schaffer, Columbia University

Date & Time

Please join us for a CCN Seminar on February 9th at 10:30AM with guest speaker, Evan Schaffer, Columbia University who will be presenting, “Order and disorder across spatial and temporal scales and across brains.”

Abstract: What are the spatial and temporal scales over which information is encoded in neuronal activity? In different experimental settings, the activity of either highly structured localized circuits or unstructured distributed neural populations has been linked to specific computations and behaviors. In the first half of my talk, I will present a theoretical analysis of unstructured neural populations with apparently random connectivity, using the olfactory system as an example. Random connectivity implies that odor representations are not only disorganized but different across individuals. How can individuals nevertheless agree on the qualities of an odor? I will show that these different representations can nevertheless support consistent agreement about odor quality across a range of odors after only a single shared experience.  In the second half of my talk, I will describe progress in understanding neural dynamics on different scales, from highly localized circuits to brain-wide activity encompassing large populations of neurons. Identifying which scales are relevant for a given function is usually not possible because observing the brain at many scales simultaneously is challenging. Fortunately, in the fly, much is known about the function of small groups of genetically identified neurons and imaging the entire brain at high resolution is tractable. We imaged the fly brain with high spatiotemporal resolution while flies engaged in a variety of behaviors. This revealed neural representations of behavior on multiple spatial and temporal scales. The activity of most neurons across the brain correlated with running over timescales that ranged from seconds to a minute.  However, residual activity not directly correlated with behavior was on average similar in magnitude to behavior-related activity. These residual dynamics were high dimensional and reflected the activity of small clusters of spatially organized neurons that likely correspond to genetically defined cell types.  These clusters participate in the global dynamics, indicating that neural activity reflects a combination of local and broadly distributed components. Thus, global patterns may inform local computation and in turn, local computations may influence global patterns. This suggests that small circuits with highly specified functions are provided with knowledge of the larger context in which they operate. Moreover, our observations imply that large-scale dynamics are not only fundamental to unstructured and distributed systems but also a component of highly structured and specialized circuits.

Onsite location will be in the GDF Auditorium at 160 5th Avenue. Zoom credentials can be found in the calendar invite for anyone tuning in remotely. 

Advancing Research in Basic Science and MathematicsSubscribe to Flatiron Institute announcements and other foundation updates