Figure 3 | Scientific Reports

Figure 3

From: E-I balance emerges naturally from continuous Hebbian learning in autonomous neural networks

Figure 3

The superimposed 10 sec traces of the membrane potentials of a network of N = 400 neurons and a relative fraction of excitatory to inhibitory neurons of 80:20. Left: After a simulation of 3600 sec for the same SOPBN considered in Figs. 2 and 4. Right: For a network with only short-term synaptic plasticity, namely with ε b  = 0 = ε w . Note that the synaptic weights are in this case as drawn from the initial distribution, which is balanced with means 7.5 (−30.0) and widths 0.375 (1.5) for excitatory and respectively for inhibitory synapses. Additionally turning off short-term synaptic plasticity changes the irregular state only quantitatively.

Back to article page