homehome Home chatchat Notifications


The surprising link between your pupils and how your brain stores memories at night

In the dim glow of the lab, mice slept peacefully like every other night. But this was no ordinary evening slumber. Scientists had mounted special equipment, including tiny cameras that recorded the movement of the rodents’ pupils while they slept. This is how researchers at Cornell University discovered that these fleeting moments of rest are […]

Tibi Puiu
January 2, 2025 @ 7:29 pm

share Share

Illustration by Midjourney.

In the dim glow of the lab, mice slept peacefully like every other night. But this was no ordinary evening slumber. Scientists had mounted special equipment, including tiny cameras that recorded the movement of the rodents’ pupils while they slept. This is how researchers at Cornell University discovered that these fleeting moments of rest are not just about recovery—they are the brain’s theater of memory consolidation, replaying the past and safeguarding the future.

The researchers found a surprising link between pupil size during sleep and how memories are processed. Their findings may offer clues to improving memory in humans and even inspire advances in artificial intelligence.

The eyes never lie (not even when you sleep)

Memory is a fragile, fleeting thing. It allows us to hold onto the faces of loved ones, navigate familiar streets, and master new skills. But how does the brain keep old memories intact while making room for new ones? Scientists have puzzled over this for decades, particularly the phenomenon known as “catastrophic forgetting,” where new information overwrites old knowledge.

“Non-REM sleep is when the actual memory consolidation happens, and these moments are very, very short periods of time undetectable by humans, like 100 milliseconds,” explained Azahara Oliva, an assistant professor at Cornell and co-lead of the study.

Oliva and her team taught mice a series of tasks—navigating mazes, collecting treats—and then monitored their brains and pupils as they slept. They noticed a pattern: during a specific substage of non-REM sleep, the mice’s pupils contracted, signaling the consolidation of newly learned tasks. Later, as the pupils dilated, older memories were revisited and reinforced.

“It’s like new learning, old knowledge, new learning, old knowledge, and that is fluctuating slowly throughout the sleep,” Oliva said.

Sleep is divided into two main stages: REM (rapid eye movement) and non-REM. Non-REM sleep is the body’s repair shop, where breathing slows, muscles relax, and memories are quietly consolidated in the deeper phases. It’s a time of restoration, both physical and mental. Then comes REM sleep, a more mysterious phase marked by darting eyes beneath closed lids, heightened brain activity, and vivid dreams.

The research reveals a previously unknown micro-structure of non-REM sleep, where the brain alternates between processing new and old memories. By disrupting sleep at critical moments, the researchers found that the mice struggled to recall either recent tasks or older ones, depending on the timing of the disruption.

Antonio Fernandez-Ruiz, co-lead author of the study, claims “the brain appears to have an intermediate timescale to organize memory processing during sleep, preventing interference between the new and the old.”

These findings challenge the traditional understanding of non-REM sleep as a uniform state. Instead, the researchers uncovered a dynamic process, where the brain cycles through distinct phases of memory work, reflected in the rhythmic changes of pupil size.

Implications for Humans and Beyond

While the study focused on mice, the implications for humans are tantalizing. Memory problems plague millions, from students cramming for exams to patients with neurodegenerative diseases. Understanding the brain’s natural system for memory segregation could pave the way for new treatments.

Moreover, the research holds promise for artificial intelligence. Neural networks, the backbone of AI systems, often face challenges akin to catastrophic forgetting. Insights from how the brain protects and consolidates memories might inspire more robust machine learning algorithms.

For now, the humble mouse and its shrinking pupils are the stars of a story that reshapes our understanding of sleep and memory. “We are just beginning to understand the incredible complexity of what happens when we close our eyes,” Oliva said. “There’s so much more to discover.”

The findings were reported in the journal Nature.

share Share

Japan is using AI to combat anime and manga piracy

Japan’s anime industry is a multibillion-dollar cultural powerhouse, but piracy poses a massive challenge.

Underwater volcano off Oregon coast on the verge of eruption in 2025

Scientists anticipate the submarine volcano will erupt before the end of 2025.

Earth Might Have Had a Ring System Like Saturn Millions of Years Ago

The ring might have acted like a giant sunshade, causing a cooling effect that might have unleashed an ice age.

America's oldest tombstone journeyed across the Atlantic and honored a knight

An old tombstone from Jamestown made a full transatlantic journey — according to its fossils.

Godfather of AI says there's a 10-20% chance AI wipes out humanity in 30 years

AI could bring an industrial revolution-level change, but at what cost?

Crafty magpies use anti-bird spikes to build their nests

It doesn’t get any crazier than this, the researchers said

Ancient Egyptians Might Have Used Water-Powered Hydraulics to Build First Grand Pyramid

Ancient Egyptians might have used hydraulic systems in pyramid construction.

The pair of jeans that sent the chess world in turmoil

Magnus Carlsen wore jeans to a chess tournament. Now the entire sport is boiling over.

Ants outperform humans at group puzzle-solving activity

Ants may have tiny brains, but when it comes to teamwork, they pack a mighty punch.

Geneticists have finally solved the mystery of Garfield’s orange coat

Two new studies have revealed why some cats are orange – an enduring enigma of genetics, until now.