Sometimes the Improvement Happens After Practice Stops

In one version of a visual-learning experiment, people improved over several days without practicing the task in between.

That sentence needs some boundaries before it becomes bad study advice.

The task was a laboratory test of visual perception, not algebra, vocabulary, piano, or basketball. Everyone had an initial learning session. “No practice” meant no additional training or reminder trials between that first session and the later test. The experiments were small. This is evidence about a particular kind of perceptual learning, not a discovery that repetition has become unnecessary.

The thing Ylka Kolken and colleagues changed was the attentional burden of the original task.

Participants performed a texture-discrimination task that required them to identify a briefly presented peripheral visual target. In the standard version they also had to perform a separate task at fixation in the center of the screen. Attention was divided.

After the initial session, some participants received full additional training. Others got only a handful of reminder trials on the intervening days. A third group received neither. In the dual-task condition, the reminders produced learning comparable to the full training sessions; the no-practice group showed little improvement.

Earlier work had already suggested that very brief memory reactivation can sometimes preserve or extend perceptual learning. The new experiment asked why those reminders help.

Then the researchers removed the central task.

What changed when the first session had full attention

With attention focused on the peripheral target alone, performance improved enough that the researchers had to make the discrimination harder. They then repeated the comparison between full training, brief reminders, and no intervening practice.

All three groups improved by the later test. Even participants who completed the initial session and then received no more task exposure showed an offline gain comparable to the groups that practiced or received reminders.

The result points toward the attentional quality of the first session as a boundary condition. Under divided attention, later reactivation appeared to matter. Under full attention, improvement emerged across the following days without it.

The authors call this multi-day latent consolidation. Performance changed after training had ended, with no new trials supplying additional task experience.

That does not mean nothing was happening between sessions. Memory and perceptual learning can continue to stabilize offline, including across periods of sleep. What is unusual here is the behavioral result: in the single-task condition, further practice was not necessary for the later improvement the researchers measured.

The study does not travel very far on its own. It used a narrow visual skill, the experimental groups were small, and the authors report limitations in the eye-tracking data for the single-task condition. Nothing here shows that someone can study a complex subject once and then wait for mastery to arrive.

It does identify a useful question for further work. Once a representation has been established under good attentional conditions, what is another practice session actually adding?

In this texture-discrimination task, sometimes the measurable improvement came later anyway.

Sources

• Ylka Kolken, Mark J. Roberts, Parisa Naseri, Federico De Martino, Nitzan Censor, and Peter De Weerd, “No practice and still perfect: Different routes for promoting offline gains in perceptual learning,” iScience 29 (11), 117839 (2026): https://doi.org/10.1016/j.isci.2026.117839
• Ylka Kolken et al., “No Practice and Still Perfect: Multi-Day Latent Consolidation in Perceptual Learning,” preprint version: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5991006
• “Brief memory reactivations enable generalization of offline visual perceptual learning mechanisms,” Scientific Reports (2025): https://doi.org/10.1038/s41598-025-06564-y