Skip to main content
. 2021 Feb 16;15:612041. doi: 10.3389/fncom.2021.612041

Figure 7.

Figure 7

(A) The spike trains of the whole neural network during R-STDP training process. Each blue dot indicates that the neurons corresponding to the vertical axis discharge at the time node corresponding to the horizontal axis. As for the number of neurons, WMC is composed of No. 1-979 neurons, No. 980–997 are motor neurons and RLC consists of No. 998 1639 neurons, respectively. (B) The weights distribution of Reinforcement Learning Circuit after implement R-STDP learning, which contains the supra-regular grammars (Mirror rule in this figure).