“…Our work focuses on one aspect of vision-the encoding of visual features; an exciting future direction is to combine encoding models with models of trial-to-trial variability [86,87,88,89,90], adaptation [91], spatial attention [92,93,94,95], arousal [96,97], visual crowding [98,99], saccade planning [100,101,102], perceptual learning [103], among other functions of V4 [27]. Overall, our compact models demonstrate that the task-driven computational models of the visual cortex are needlessly large.…”