Title: Memory encoding and continuous attractors in recurrent networks
Speaker: Professor Vladimir Itskov
Speaker Info: University of Nebraska
Special Note: More current information may be available at PlanItPurple
The hippocampus is often thought of as a “Swiss knife” of the brain. It is implicated in learning and memory, but it has also been found to be critical for spatial navigation. While some hippocampal functions have been successfully modeled using continuous attractor networks, which are examples of spatially structured networks, other functions such as associative memory encoding appear to require a different synaptic organization. How can the various functions of the hippocampus be accomplished by the same network?Date: Monday, February 6, 2012
In my talk I will first discuss spatially structured networks, i.e., networks of neurons where the synaptic efficacies are organized by the cells' positions in a "feature space" that reflects their coding properties. Next I will propose a simple mechanism by which such networks can naturally arise and be maintained in the hippocampus. I will then describe a framework that allows associative memories to be encoded in a recurrent network via "perturbations" of a spatially structured network, and expose the mathematical approaches for tackling this problem.