Solvable neural network model for input-output associations: Optimal recall at the onset of chaos
Solvable neural network model for input-output associations: Optimal recall at the onset of chaos
In neural information processing, inputs modulate neural dynamics to generate desired outputs. To unravel the dynamics and underlying neural connectivity enabling such input-output association, we propose an exactly solvable neural-network model with a connectivity matrix explicitly consisting of inputs and required outputs. An analytic form of the response under the …