Penn Arts & Sciences Logo

CNI Speaker Series: Alexei Koulakov

Monday, January 30, 2017 - 12:15pm

Barchi Library (140 John Morgan Building)

Alexei Koulakov
Professor of Neuroscience
Cold Spring Harbor Laboratory

Neural relativity principle

In the olfactory system, odor percepts retain their identity despite substantial variations in concentration, timing, and background. I will describe a strategy for encoding intensity-invariant stimulus identity that is based on representing relative rather than absolute values of the stimulus features. Because, in this scheme, stimulus identity depends on relative amplitudes of stimulus features, identity becomes invariant with respect to variations in intensity and non-linearities of neuronal responses. In the olfactory system, stimulus identity can be represented by the identities of the p strongest responding odorant receptor types out of a species dependent complement. I will show that this information is sufficient to recover sparse stimuli (odorants) via elastic net loss minimization. This minimization has to be performed under a set of constraints, including the identities of the strongest responding receptors. This problem can be solved using dual representations, in which odorant identity is encoded by a set of Lagrange multipliers. I will show that the dual problem, in turn, can be solved by a neural network whose Lyapunov function represents the dual Lagrangian. Thus, I will present a model for the olfactory cortex that computes concentration-invariant odorant identity by implementing dual representations with the sparse activities of individual neurons representing the Lagrange multipliers. Sparseness of neural activities emerges in this approach from the KKT theorem.


A pizza lunch will be served.