Figure 3.

Description of the three learning rules. (A) Time course of the variables involved in the optimal model. $\widetilde{\Delta w}$ denotes the cumulative weight change. (B) Schematic representation of the phenomenological models of STDP used in this paper. Each presynaptic spike yields LTD proportionally to o₁ (blue trace) in both models (pair and triplet). In the pair model, postsynaptic spikes evoke LTP proportionally to r_j (green trace), while in the triplet model r_j is combined with an additional postsynaptic trace o₂ (red).