Figure 11. Five-neuron ensemble with a hidden positive feedback network.

(A) Five neurons interacting with one other were simulated, but only neurons 1, 2, and 3 were treated as spike trains that were simultaneously recorded together. The activities of neurons 4 and 5 were used to create a hidden positive feedback network with neuron 1, but were treated as neurons outside the receptive field of an electrode: that is, their activities were not used as covariates to create the connectivity maps. (B) The true causal connectivity map between neurons 1, 2 and 3 is obtained from (A).