Conscious perception may be less steady than it feels. According to a preprint posted on bioRxiv on October 16, 2026, activity in monkey cortex alternates about 15 times per second between a consciously seen image and a suppressed one. The work comes from Janis K. Hesse, Doris Y. Tsao and colleagues, and it has not been peer reviewed. This article is based on the preprint’s abstract and the summary and introduction text we could read.
How the study tested conscious perception
The team used Neuropixels probes to record activity across several face patches in inferotemporal and prefrontal cortex. The task was binocular flash suppression. In this set-up, one image is consciously perceived while the other, although continuously present, is suppressed from awareness.
The authors say a large set of face images let them reconstruct which specific faces the neural population was representing. They also say the probes let them track changes within individual trials, at a resolution that single-neuron recordings cannot offer.
What the researchers report about seen and suppressed images
The headline result is a rhythm. According to the abstract, during a constant stimulus and a constant percept, the brain rapidly alternates, about 15 times per second, between “conscious” and “suppressed” micro-epochs.
During conscious micro-epochs, the authors say a pure signal of conscious perception was present, which allowed accurate reconstruction of the conscious image from neural activity. In suppressed micro-epochs, this reversed: the signal tracked the suppressed image instead. At the level of single cells, neurons responded selectively to the conscious stimulus in the first kind of epoch and to the suppressed stimulus in the second.
The two kinds of epoch could still be told apart, with feature tuning amplified during conscious micro-epochs. Recordings in prefrontal cortex gave qualitatively similar results. The authors also describe a comparison in which activity averaged over the whole period, without separating the states, produced reconstructions that matched neither face but looked like a blend of the two.
Why this matters for the study of consciousness
The abstract notes that earlier studies found high-level visual cortex to be only weakly modulated by conscious perception, and that a pure signal of conscious perception had not been found anywhere in the brain. Their time-averaged comparison hints that averaging over time may hide a faster alternation. From the results, the authors conclude that conscious access requires temporal selection beyond localization, meaning timing matters, not only where in the brain activity occurs.
For other recent brain-related research on our site, see our report on a fly brain wiring map with 166,700 neurons, our look at a light-powered brain that learned 81 light patterns, and our explainer on neuroscience and the mysteries of the mind.
Limits of this conscious perception preprint
The first limit is status. A bioRxiv preprint is shared before peer review, so experts have not yet checked the methods or the statistics, and no independent group has reported repeating it. Everything here is the authors’ own account.
Second, the study was done in macaque monkeys and used face images. The abstract does not say whether the same pattern would appear in humans or with other kinds of stimuli. Third, the 15-per-second alternation is described in neural activity during a stable percept. It does not mean an animal, or a person, sees the image flicker 15 times a second.
Finally, we did not have access to the full methods, so details such as the number of animals and the exact statistical tests are not covered here.
My view on conscious perception research
As a writer, I think this is an interesting idea that deserves a careful reading. If the same neurons really do carry the seen and the unseen image in quick turns, it could explain why earlier experiments found weak signals. But “a pure signal of consciousness” is a strong phrase, and I would want to see other laboratories reproduce it, ideally with different stimuli and in other species, before treating it as settled. For now it is a promising hypothesis backed by new recordings, not a solved problem.
FAQ: conscious perception and the 15-per-second finding
What did the preprint find?
According to the authors, activity in macaque inferotemporal and prefrontal cortex alternates about 15 times per second between representing the consciously perceived image and the suppressed image, even when the stimulus and percept stay constant.
Has the study been peer reviewed?
No. It was posted on bioRxiv on October 16, 2026, and the findings have not yet been through peer review or independent replication.
Does this mean we see the world flicker?
Not according to the abstract. The alternation is measured in neural activity while the percept stays constant, so it does not describe a visible flicker.