In partnership with

Most anti-aging devices are expensive placebos. This one isn't.

Red light therapy sounds like wellness bullsh*t. We get it.

But Celluma is the only LED brand with 5 FDA clearances, not just for wrinkles, but also pain relief, hair growth, body contouring & more. Clinical trials showed 80% improvement in skin texture and a 66% reduction in wrinkles after 4 weeks of consistent use.

Most LED devices don't have the power density or wavelength precision to actually do anything. Celluma does. With proprietary processor-driven algorithms that deliver proven results, plus a patented design that shapes around your face, scalp, or any body part. No needles, no downtime, no Botox appointments.

The only catch? Consistent use. This isn't a one-and-done fix. It's mitochondrial support for your skin & entire body, a fancy way of saying it helps your cells do their job.

If you're already spending money on anti-aging stuff anyway, at least try something with actual clinical backing.

Get up to $300 off Celluma

_{Results vary. Consistency required. FDA-cleared for specific indications.}

Introduction

For decades, neuroscience has largely assumed that learning improves through repetition. The dominant framework in reinforcement learning suggests that each experience slightly updates the brain’s prediction of future rewards, gradually strengthening associations through repeated trials.

A new study published in Nature Neuroscience challenges that assumption. Researchers found that the speed at which animals learn associations is not determined by the number of repetitions they experience, but by the time between rewards. When rewards were spaced further apart, animals learned dramatically faster per experience, even though the total learning time remained essentially unchanged.

In the study, researchers trained mice to associate a brief auditory tone with a sucrose reward while systematically changing the interval between rewards. By extending the interval between rewards from 60 seconds to 600 seconds, the animals required about ten times fewer learning trials, yet reached the same learning point in roughly the same total amount of time.

What the Research Showed

Across multiple experimental conditions, the same pattern consistently appeared: learning scaled with the time between rewards rather than the number of learning experiences.

In one comparison, mice receiving rewards every 60 seconds required roughly 94 trials to learn the cue–reward association, while mice receiving rewards every 600 seconds learned the same association in only about 8.8 trials.

Despite this dramatic difference in repetitions, the total time required to learn remained nearly identical, roughly six thousand seconds in both conditions. In other words, removing nine out of ten learning trials did not meaningfully slow learning.

When researchers examined multiple reward intervals ranging from 30 seconds to 600 seconds, they found a near-perfect mathematical relationship between reward spacing and learning rate. As the time between rewards increased, the number of trials required for learning decreased in almost exact proportion.

The consistency of this pattern suggested that the brain may be following a simple rule: the less frequently rewards occur, the more learning occurs from each one.

Mechanisms & Neuroscience

Dopamine and Reward Prediction

Dopamine neurons play a central role in reinforcement learning. For decades, neuroscientists have believed that these neurons encode what is known as a reward prediction error, a signal that indicates the difference between expected and received rewards.

When a reward occurs unexpectedly, dopamine neurons fire strongly. As the brain learns which cues predict rewards, this dopamine signal gradually shifts from the reward itself to the predictive cue.

In this study, researchers measured dopamine activity in the nucleus accumbens, a key component of the brain’s reward system. They found that dopamine learning signals followed the same scaling rule as behavior: when rewards were spaced further apart, dopamine responses to the predictive cue emerged in far fewer trials.

This indicates that the change in learning speed was not merely behavioral, it was reflected directly in the brain’s core reward-learning circuitry.

The Mesolimbic Learning Circuit

The dopamine signals measured in this study originate within the mesolimbic reward circuit, one of the brain’s most important systems for learning from experience.

In this circuit, dopamine neurons in the ventral tegmental area (VTA) project to the nucleus accumbens, where dopamine release influences synaptic plasticity. These signals help the brain strengthen neural connections linking environmental cues to meaningful outcomes.

Over time, cues that reliably precede rewards gain motivational significance. This process allows the brain to learn which signals in the environment predict valuable outcomes and adjust behavior accordingly.

Because dopamine-driven plasticity shapes how strongly these cue–reward associations form, changes in dopamine signaling can directly influence how quickly learning occurs.

A New Learning Algorithm?

The study also explored whether existing computational models of learning could explain the observed results.

Most traditional models of reinforcement learning assume that learning progresses on a trial-by-trial basis, meaning each experience contributes a small update to the brain’s predictions. These models predict that more repetitions should lead to faster learning.

However, the experimental data did not follow this pattern.

Instead, the researchers found that the results were better explained by a retrospective causal learning model, in which the brain updates associations based on how often a cue precedes a reward over time. In this framework, when rewards occur less frequently, each reward provides more information about the events that caused it.

As a result, the brain may update its internal model more strongly from each reward when rewards are rare.

Practical Applications for Brain Health

These findings offer a new perspective on how the brain extracts meaningful information from experience.

First, they help explain a long-observed phenomenon in psychology known as the spacing effect, where learning and memory often improve when experiences are distributed across time rather than concentrated together.

Second, the results suggest that the brain may prioritize informational value over repetition. When outcomes are rare, each reward provides a stronger signal about what caused it, allowing the brain to update associations more efficiently.

Finally, this mechanism may also help explain why certain forms of reinforcement, such as intermittent rewards, can produce powerful behavioral conditioning. When rewards are unpredictable or infrequent, the brain may treat each one as a highly informative event.

The Bottom Line

Learning may not simply depend on how many times an experience occurs.

Instead, the brain appears to adjust how strongly it learns from each outcome depending on how frequently rewards occur in time. When rewards are rare, each one carries more informational value, allowing the brain’s dopamine system to update associations more powerfully.

Reference

Duration between rewards controls the rate of behavioral and dopaminergic learning
Nature Neuroscience
DOI: 10.1038/s41593-026-02206-2