10.1371/journal.pcbi.1002494.g003 Michiel W. H. Remme Michiel W. H. Remme Wytse J. Wadman Wytse J. Wadman Adapting to a persistent change in external drive to a recurrent network that includes two types of inhibitory cells. Public Library of Science 2013 persistent recurrent includes types inhibitory 2013-02-20 05:33:58 Figure https://plos.figshare.com/articles/figure/_Adapting_to_a_persistent_change_in_external_drive_to_a_recurrent_network_that_includes_two_types_of_inhibitory_cells_/313279 <p>A: Response of a recurrent network with two types of i-cells: half of the i-cells show HSE while the other half does not show HSE at the depicted time scale. All cells are perturbed by a 1.2 kHz step increase in the external drive (top trace). Top panel shows the population mean of the instantaneous firing rates of the e-cells (black), the adapting i-cells (red), and the non-adapting i-cells (magenta), computed for 1 second bins. Bottom panel shows the population mean of the three populations. The adaptation time scale ratio of the (adapting) i-cells to the e-cells is 0.5. The compound external drive to all cells before the input increase is 1.2 kHz. B: Ratio of adaptation time scales that is required for adaptation dynamics to be stable when the fraction of all i-cells that adapt is varied from 10% to 100%. The network is perturbed from its adapted state (to a 1.2 kHz external drive) by a 0.3 kHz step increase of the external drive to all cells. The same criterium for stability of the adaptation dynamics is used as in <a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1002494#pcbi-1002494-g002" target="_blank">figure 2</a>.</p>