Duke Physics Colloquium: Storing and retrieving memories in models of neuronal networks

"Storing and retrieving memories in models of neuronal networks" - Memories are thought to be stored in brain networks thanks to modifications of synaptic connectivity between neurons. Mathematical models of synaptic plasticity (sometimes called `synaptic plasticity rules' or `learning rules') capture experimental data on plasticity with increasing accuracy, but it is still unclear how realistic synaptic plasticity rules shape network dynamics and information storage. In this talk, I will first review approaches for inferring learning rules from neurophysiological data. I will describe in particular a new approach in inferring the
learning rules from in vivo electrophysiological data, using experiments that compare the statistics of responses of neurons to sets of novel and familiar stimuli. I will then focus on how the inferred learning rules shape the dynamics of networks, leading to a diversity of potential dynamics that allow the network to retrieve the stored information (fixed point attractors, chaotic attractors, or transient sequential activity). Finally, I will show that learning rules inferred from data are close to maximizing information storage. Faculty host: Anselm Vossen. Refreshments will be available before the event in room 130.