Figure 10. Stable learning in a topographically organized recurrent two-dimensional neuronal network with in vivo-like spontaneous activity. A, Time evolution of the unstimulated population (gray curve) and stimulated population by an external drive within a circular area (black curve). The mean afferent synaptic weights for each neuron in the network are displayed as snapshots in the course of time above the graph. The mean afferent weights fed by neurons inside the stimulated region, outside the stimulated region, and from all neurons are plotted in the first, second, and third rows, respectively. B, From left to right, The columns labeled by circled numbers correspond, respectively, to the stable spontaneous regime (1), the end of the stimulation epoch (2), and the relaxation ongoing regime 1000 s after the stimulation (3). Network snapshots of the mean afferent and efferent synaptic weight per neuron are plotted in the two first rows for mSTDP. In the last two rows, the mean firing rates for each neuron in the network are plotted for the mSTDP and the weight-dependent STDP rules. Circled numbers at the top of each column refer to the same phases described in A and B. C, Network firing statistics before, during, and after stimulation: top, firing rate distribution; middle, ISI CV; bottom, mean Pearson's coefficient computed over 2000 distinct neuron pairs.