Fig. 1

The input and response of a suboptimal network. The top three lines show the input signals, colour coded for the order in which they appear (to facilitate the analysis of the network activity). The bottom line shows the voltage of the output neuron (the activity of the interaneurons is not shown). During artificial evolution, the spikes after the input sequence “A B C” are rewarded (green rectangles) and any other spikes (red circles) are penalized (such spikes occur very rarely in the response of the best networks at the end of the successful evolutionary runs)