Back

High-fidelity, but hypometric spatial localization of afterimages across saccades

Schweitzer, R.; Seel, T.; Raisch, J.; Rolfs, M.

2025-05-09 neuroscience
10.1101/2025.05.04.652081 bioRxiv
Show abstract

Humans typically perceive their visual world as stable and continuous, despite frequent shifts of the retinotopic reference frame caused by saccades. This visual stability is paralleled by afterimage movement across saccades: Although retinotopically stable, afterimages appear to move in egocentric space wherever the eye moves. To investigate the mechanisms underlying this phenomenon, we tasked human observers to localize afterimages relative to briefly flashed probes in complete darkness. This psychophysical tracking of afterimages was accompanied by eye tracking, allowing us to fit a dedicated computational model to accurately predict afterimage movement based on the size of eye movements. The gain of afterimage movement was significantly hypometric, remained unaffected by post-saccadic visual feedback and saccadic adaptation, and was inversely related to saccade gain. These findings suggest that afterimage movement is driven by efference-based, feedforward prediction of visual consequences of saccades and demonstrate the potential of the afterimage-tracking technique for studying visual stability. SignificanceIn low-light environments, brief high-intensity visual stimulation can induce long-lasting retinal afterimages. When observers then make eye movements to explore their visual environment, these afterimages - albeit fixed in the retinotopic frame of reference - appear to move in egocentric space wherever the eye moves. Even though this phenomenon has been known for centuries, the underlying computations remained unexplained. Tracking eye and afterimage positions simultaneously, we found that perceived afterimage position was accurately predicted by eye position across a variety of visuomotor conditions, whereby the eye movements size was however systematically underestimated by the visual system. Considering a parsimonious model of visual localization, afterimage movement can be understood as a consequence of feedforward predictions of the visual consequences of impending eye movements.

Published in Science Advances · training set

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.