Memory consolidation in a neuronal model.

<p><b>a.</b> The <i>effective</i> learning rate of a downstream synapse depends on the transfer process itself. Increasing the number of transfer repetitions increases this rate leading to faster learning and faster forgetting. Shown is SNR of each of the first two stages. Symbols are averages of ten simulations, lines are from the mean-field model, see <a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1003146#s4" target="_blank"><i>Methods</i></a>. Here , , , and which gives and when . <b>b.</b> The neuronal model is well described by the mean-field synaptic model. There are 10 stages, each with all-to-all connected neurons. Parameters are chosen such that transfer rates are . The solid line is for a in the mean-field model. Shown is the combined SNR for all 10 stages.