Skip to main content
. 2022 Dec 8;11:e77603. doi: 10.7554/eLife.77603

Figure 11. Simulation performance with and without Purkinje cell (PC) feedback to molecular layer interneurons (MLIs).

Figure 11.

(a) Examples of virtual conditioned responses (CRs) from the last session of training at interstimulus interval (ISI) 200 for simulations without feedback from PCs to MLIs (conditioned stimulus [CS] is shown in red) and with feedback collaterals (CS is shown in black). (b) Acquisition curves for simulations constructed with (black) and without (red) feedback from PCs to PC-MLIs, plotting the amplitude of virtual CRs as a function of sessions of training for the three ISIs indicated (108 trials per session). (c) Average virtual CR amplitude on the final session of training is shown for training using 11 ISIs ranging from 150 to 1000 ms; black bars indicate results from simulations with PC to PC-MLI feedback and red bars depict data from simulations lacking feedback. Simulations with PC feedback performed considerably better at relatively long and short ISIs, while for intermediate ISIs there was less effect on performance.