HMN 2026: How Gazing longer at something contributes to memory encoding

While humans are observing their surroundings, their eyes tend to rapidly shift between different objects, people and details that catch their attention, pausing briefly on each one. In psychology, prolonged pauses on specific items are known as fixations.

Many past studies exploring aspects of human attention and information processing relied on eye-tracking technologies. Most of these studies assumed that prolonged fixations are a sign of focused attention and of the brain processing specific visual information.

Yet researchers at Osnabrück University, the Max Planck Institute for Human Cognitive and Brain Sciences and Justus Liebig University Giessen recently gathered new evidence indicating that this is not always the case. Their paper, published in Nature Neuroscience, suggests that the duration of fixations is not always related to visual processing demands but might instead be a sign of the brain’s efforts to encode visual information and form new memories.

“Where we look is one of the richest windows we have into the mind and the processes that govern it,” Philip Sulewski, first author of the paper, told Medical Xpress.

“Yet for one basic feature of looking, how long we hold a fixation before moving on, we still have remarkably little idea what the brain is computing during a long fixation as opposed to a short one. Yet the differences in duration are stark: Some fixations last under 150 ms, others more than 500 ms. For a brain that processes information very quickly, that is a long time to linger.”

Studying the brain while people observe natural scenes

Sulewski and his colleagues wanted to better understand what the brain is engaged in while humans are staring at objects for longer periods. To do this, they carried out a study involving five participants who took part in several experimental sessions, with about one week between each session.

During these sessions, the participants were shown thousands of images capturing natural scenes. While they were looking at each of the images, the researchers recorded their eye movements using eye-tracking technology and monitored their brain activity using a noninvasive neuroimaging technique called magnetoencephalography (MEG).

MEG works by measuring faint magnetic fields produced by neurons when they fire. The team later analyzed MEG scans in conjunction with the gaze-related information they collected to better understand what areas of the brain were active when participants’ eyes were lingering on specific details in the images.

“Over the experimental sessions, participants viewed more than 4,000 natural scenes, far more than is usual, so that we could capture the variety of scenes we meet in everyday life and study fixation behavior at scale (in total, more than 235,000 individual fixations),” Sulewski said.

“We recorded brain activity with MEG alongside eye tracking. We then derived, for each fixation, an estimate of how difficult the fixated content was to process, together with two separate estimates of downstream memory processing. This let us test, fixation by fixation, how recognition difficulty and memory processing relate to how long the eyes stayed in place.”

Using artificial intelligence (AI) models, the researchers derived estimates of how difficult different segments of images were to classify and how memorable they were.

Interestingly, the results of their analyses suggested that the segments of images that were harder to classify were typically those that people looked at for shorter periods. In contrast, participants appeared to glance longer at regions in scenes that were predicted to be more memorable.

Rethinking the meaning of prolonged gaze

Collectively, the findings gathered by Sulewski and his colleagues challenge the assumption that longer fixations reflect extensive visual processing. Instead, they suggest that looking at specific items for a longer time supports the formation of memories.

“There is a strong assumption, inherited from the dominant tradition of object-recognition research using fixed gaze and randomly presented stimuli, that longer reaction times reflect harder visual processing,” Sulewski said.

“Our results indicate that this intuition does not generalize to fixation timing in more naturalistic conditions. Instead, we find that fixation timing is more closely tied to memory encoding than to visual processing difficulty: The brain appears to linger to commit information to memory rather than because it is still working to recognize what it sees.”

This study could soon inspire more research aimed at validating the idea that long fixations are linked to memory encoding. Sulewski and his co-authors are also planning new studies designed to further explore the link between memory and gaze.

“In the longer term, a reliable link between memory and gaze could in principle inform diagnostics for memory impairments, though I would stress that this is speculative at this stage,” Sulewski added. “Meanwhile, we will continue to work out how memory and gaze interact at scale under naturalistic conditions and try to develop better computational models of that process.”

The content is provided for information purposes only.