Figure 5. Saccade-triggered responses are less bimodal during rivalry.

(a) Single-trial responses during saccades averaged across simultaneously recorded ML neurons from the same session as in Figure 4b that were significantly face-selective for both physical and perceptual conditions. Individual neuron responses were normalized to make the mean object response −1 and the mean face response +1. Rows of each plot correspond to response time courses to individual saccades, aligned to saccade onset, and sorted by average response during 0–400 ms after saccade onset. Top: Physical condition. Bottom: Perceptual condition. Left, middle, and right columns correspond to saccades during object epochs, face epochs, and across both, respectively. The difference between perceptual and physical conditions in the third column shows that this difference cannot be simply attributed to mislabeling of perceptual state by the no-report paradigm. (b) Histograms of saccade-aligned responses averaged across a time window of 0–400 ms after saccade onset and across neurons (after normalizing as in (a)) that were significantly modulated for both physical and perceptual conditions. Top: Physical condition. Bottom: Perceptual condition. Left: Saccades for face and object plotted separately in red and blue, respectively. Middle: Saccades for either face or object epochs plotted in gray. Right: Absolute values of normalized responses plotted in light gray.

Figure 5—figure supplement 1. Lack of bimodality is a general trademark of rivalry.

(a) Trial responses in ML are less bimodal during rivalry. Histograms have same conventions as Figure 5b but instead of averaging neuron responses for individual saccades, responses are averaged across trial duration for individual trials. (b) Trial responses in AM are less bimodal during rivalry. Same conventions as in (a), but instead of the Neuropixels-like probe in ML, cells were recorded simultaneously from AM using a 32-channel S-probe.