Neuroscience Seminar Series
Vendredi 7 juin 2019, 11:30, R229 (2e étage), Centre Universitaire des Saints Pères, 45 rue des Saints Pères, 75006 Paris
Brent Doiron, PhD, Professor, Department of Mathematics University of Pittsburgh ,USA
Neuronal responses are notoriously variable, with sizable trial-to-trial and dynamics fluctuations in spiking activity that are shared across large populations of neurons. Furthermore, the degree and population structure of this variability is malleable, depending on host of stimulus and cognitive factors. I will present modelling and theoretical work that uncovers how spatially extended cortical circuits with large excitation that is balanced by an opposing inhibition can capture low dimensional shared variability reported in many population recording studies. The spiking variability in our model is also easily quenched through a top-down signal to inhibitory neurons, matching experimental results in spatial attention discrimination task. Finally, we explore how our circuit based manipulation of neuronal variability affects information flow as it propagates across cortical areas.