Fig. 1

Illustration of a sparse recurrent neural network with five nodes in a linear array. All the nodes exhibits self-excitation, represented by \(+\alpha\) above the central node, while lateral inhibition between neighboring nodes is given by \(\beta _1\) and inhibition between nodes separated by another is shown as \(\beta _2\) (\(\beta _2 < \beta _1\)). Transient input is shown as \(I_i\) below each node.