A Reinforcement Learning Framework for Spiking Networks with Dynamic Synapses

Figure 4

Evolution of the trained parameters: 𝜏_fac in (a), 𝜏_rec in (b), and 𝑈_SE in (c) over time. In all subfigures, samples from an excitatory synapse (Exc.) and from an inhibitory (Inh.) one are given. For both types, the starting values are identical.

(a)

(b)

(c)