Feature selectivity in a cortical module with short-range excitation

摘要

We study analytically the stable states of a neural network model with short-range excitation and long range inhibition that codes for a one-dimensional feature. The effects of the amplitude of modulation of the interaction and their range relative to the network size on the form of the population activity profile is analyzed. We derive the conditions for the emergence of a line of attractors in the case of a uniform external input and describe its impact on the system's response to a tuned input.