Figure 23From: Interactions of multiple rhythms in a biophysical network of neurons(a) An implementation of the network topology of Fig. 22 with three copies of the network of Fig. 9(a) and an RS cell. The three “input-1” (see also Fig. 9(a)) are independent, while the “input-2” that each SI cell is receiving is a delayed version of each other. The SI→RS synapses in this model are excitatory, contrary to the fact that the SI cells are considered inhibitory. (b) A more biologically realistic version of a subcomponent of the network in (a). To avoid having a cell with both an excitatory outgoing synapse and an inhibitory autapse (like the SI cells in (a)), we may consider a pair of cells, an excitatory (E) and an inhibitory (I) one. The E cell receives the inputs and provides the output, while at the same time projecting onto the I cell, and the I cell projects back to the E cell. Assuming that the I cell fires always immediately after the E cell, the pair of reciprocal connections (E→I and I→E) function effectively as self-inhibition for the E cell, playing the same role as an inhibitory autapse would. For simplicity, in the simulation of Fig. 24 we use the version of the network shown in (a)Back to article page