Computational geometry for modeling neural populations: From visualization to simulation - Fig 10

A: compares the response for three cases: no adaptation; only current adaptation; and both current- as well as spike-based adaptation is included. B: The gain for a small sinusoidal input modulated on a background input as function of frequency, for no adaptation and AdExp with current- and spike-based adaptation. Both spectra show the dependency expected of an exponential-integrate-and-fire neuron, as the spike shape, represented by the grid at high V values is independent of w. However, the numerical difference between the cases is vast.