Appendix 2—figure 1. A trained recurrent network exhibits sequences that are less robust and less able to generalize.
(a) Schematic of the model. The cortical units receive trial-specific inputs and project to striatum. Striatal units are not recurrently connected. The corticocortical and corticostriatal weights and are set as per the text. (b) Model simulation. Upper: Inputs to model. Red trace marks initiation of trials; blue trace indicates the target time for the trial. Middle: Sample cortical units. Lower: Striatal units. (c) Means and standard deviations of best-match times as a function of target times with random weight perturbations of 2 or 5%. Open symbols denote target times for which the nRMSE exceeded 0.3 on greater than 25% of trials. (d) Best-match times for a model trained on time intervals ranging from 0.4 s to 1.2 s, and then tested from 0.1 s to 1.5 s. The colors indicate the means of the nRMSEs of the trials at each target time.