David Hansel, CNRS-Université Paris Descartes , Paris and Interdisciplinary Center for Neural Computation , Jerusalem
NBRL Seminar room (Fishbach 430) at 12:00
7 March, 2010
Balanced working memory
Persistent activity in cortex is the neural correlate of working memory (WM). In persistent activity, spike trains are highly irregular, even more than in baseline. This apparently innocuous feature challenges our current understanding of the mechanisms underlying WM.
Here we argue that in WM the PFC operates in a regime of balanced excitation and inhibition and that the observed temporal irregularity reflects this regime. We show that this implies that the interactions between PFC neurons are nonlinear. Based on this result we put forward a novel WM mechanism which draws on recent results demonstrating short-term facilitation in recurrent synapses between pyramidal cells in PFC. This model is the first to account not only for irregular persistent firing but also for the diversity of neuronal responses in PFC during WM tasks. Our work highlights the crucial role of STP in WM and more generally for computation in balanced cortical networks.