Figure 7
From: Encoding information into autonomously bursting neural network with pairs of time-delayed pulses
The dependence of RBTP on the groups of neurons that receive a probing pulse (a) RBTP vs. group used for probing pulse stimulation (#0 red is for the same group of neurons that were used for Δt training beforehand, all others are just randomly selected groups of neurons), (b) The overall increase in the mean synaptic weight of neurons which were chosen to receive the Δt = 100 ms training pulses, and (c) RBTP vs. 〈Wij〉group after training. Note in (a) that RBTPs for 10 randomly chosen groups of neurons all are not as good as that of the first recurrent burst generated by a probing pulse delivered to the group of neurons that were used for the prior Δt training.
