Fig. 4: Model of non-linear integration to generate time percept.

a Left box: stimulus-dependent input currents that follow characteristic dynamics for vibrissal and optogenetic drive, converge on vS1 neurons, giving rise to spike trains with Poisson statistics. Example vS1 PSTH is shown in the lower right corner; binned neuronal activity (solid line) in response to a vibrissal stimulus (black), as well as with photoexcitation (light blue), are simulated (dashed line with confidence intervals) by fitting the respective input currents and feeding them to an I/F. Current and I/F parameters in Methods. We simulated a 5000-neuron population based on the distribution of fitting parameters of the entire population of recorded neurons. Right box: the LI receives input spike trains from the simulated vS1 population under conditions including vibrissal (black spike train) and vibrissal plus photoexcitation (blue spike train). The integrator’s accumulated quantity is governed by the differential equation. Reading out the generated vS1 neuronal population activity with this LI, we can predict the perceptual shift created by optogenetically increasing firing rate in vS1. b Psychometric curves of the neurometric model (points). Inset: comparison of the perceptual shifts between behavioral data and generated neurometric curves for resampled (100×) neuronal and behavioral data. Bias quantified as the difference in percent of choices T2 > T1 averaged across all data points, excluding ΔT’s of 0.35 and −0.35 (for details, see Methods). c vS1 role in compressing or dilating perceived time by its sensory drive; optogenetic manipulation slows or speeds the accumulation of drive in the perceptual hourglass. The selected changes in perceived time derived from the shift in the point of subjective equality (PSE) in the averaged behavioral data. Source data are provided as a Source Data file.