Figure 7
From: Dendritic processing of spontaneous neuronal sequences for single-trial learning
Orthogonal memory formation in a Y-maze. (a) Behavioral paradigm is schematically illustrated. The animal starting from the junction successively visits three arms on a Y-maze. (b) In the initial setting of the recurrent network, a DG input triggers three firing sequences when the animal is at the junction of the Y-maze. Each branch of sequence consists of 150 neurons. (c) Spontaneous activity during the resting state is shown for the three branches of the two-compartment network model before the exploration. (d) The dendritic and somatic activities of the two-compartment network model during the first run on the Y-maze. Time evolution of inhibitory inputs to the dendritic compartments is also shown. (e) Activity of the two-compartment network model during the second run is shown. (f) Activity of two-compartment model without dendritic inhibition during the first run on the Y-maze. (g) With and without dendritic inhibition Information per spike was calculated over five simulation trials with different random seeds.
