License 1 - License 2
Mentions: This population statistic approach lends itself to quickly analyzing the response properties of population-scale dynamics of neural tissue. We use this strategy to examine the input-output relationship of the Potjans and Diesmann column model (see also ), in an attempt to uncover canonical computations  that it might implement. We find that excitatory perturbations to layers 4 (a site of primarily ”bottom-up” thalamic input in sensory areas) and 2/3 (a site of primarily ”top-down” cortical input) elicit an attenuated, additive perturbation in layer 23 activity, yet offset subtractively in their effect on layer 5 (see Figure 1). This computation might subserve, for example, an inferential update of prior experience with new sensory information. We generalize this finding by computing a linear kernel that describes the response of the column circuit to time varying stimuli.
No MeSH data available.