Fig. 7: Model predictions for stimulation experiments.

a Saccades evoked by stimulating the OC units at different delays with respect to the time at which the target position cue disappeared, while Model 4 performed the navigation task. Each line corresponds to a different trial, and color denotes the type of stimulation (see text). b Similar to A but showing the response of Model 2. c Variability in the final position of the saccade across trials as a function of the timing of the stimulation. Error bars denote ±1 standard error of the mean estimated by bootstrapping. d Comparison of the magnitude of saccade against the model’s belief about the relative target position at the time of stimulation for 4 different stimulation sites (generating rightward, leftward, downward, or upward saccades). Thin and thick circles denote the x (horizontal) and y (vertical) eye movement components, respectively. Data plotted are from all stimulation times. e Movement trajectories of the model under baseline trials (left) and during stimulated trials (middle). Right: Trajectory of Model 2 during stimulation. f Steering errors of Model 4 (solid) and Model 2 (dashed) in stimulated trials. Gray line denotes the error during baseline trials. Error bars denote ±1 standard error of the mean estimated by bootstrapping.