Fig. 3: Quantifying the spatial bias and its dynamics over time, delay, and ensembles.

a The spatial bias as a function of stimulus delay values, for a probe that appears at a location close to the ST, measured using the neurons’ sensitivities at time 100 ms after saccade onset in an example ensemble. b The spatial bias as a function of time relative to saccade onset, for the same probe and ensemble in a, measured using the neurons’ sensitivities at delay 110 ms relative to each timepoint (x-axis). c Mean spatial bias across time and delay, for 6 probe locations around the ST, averaged across all 15 ensembles (n = 447 neurons). Dashed lines indicate the range of delay values used in d. d Mean spatial bias over time, for delay 50:100 ms, for 6 probe locations around the ST, for all 15 ensembles. Shaded area represents the standard error of the mean (SEM) across ensembles. e Spatial bias over time from saccade onset, for ensembles with various distances between their neurons’ RF center and the ST (d). There are 4 ensembles with d < 11, 4 ensembles with 11 ≤ d < 14, 3 ensembles with 14 ≤ d < 17, and 4 ensembles with d ≥ 17. Shaded area represents SEM across ensembles. f Spatial bias for the 15 ensembles during the perisaccadic window (50:150 ms from saccade onset, gray bar in e), plotted against the distance between RF center and the ST for each ensemble. Error bars indicate the SEM of the bias estimate over resampling the neuronal population in each ensemble (n = 100 samples, 90% of neurons in each sample). Source data are provided as a Source Data file.