Plate Nº 40 · recorded October 10, 2026
Neuroscience & MindReported finding
Study: Under Uncertainty, Brain Holds Onto Irrelevant Visual Cues
Humans and monkeys make predictable extra errors when switching visual tasks under uncertainty, a University of Chicago team reports in Nature Neuroscience. The mechanism could help fingerprint flexibility disorders.
By Marcus Bennett3 min read553 words
In brief
- Study published in Nature Neuroscience, 2026 (DOI 10.1038/s41593-026-02430-w), led by University of Chicago co-first author Cheng Xue
- Both humans and macaque monkeys showed extra errors when switching between two visual tasks under rule uncertainty
- Mechanism, dubbed 'feature interference,' was confirmed in V1 — the first stage of visual cortex processing
- Cognitive flexibility declines with age and in Alzheimer's, Parkinson's, schizophrenia, and substance use disorders
- A related PNAS 2026 paper by Ruff, Sheets and colleagues found similar interference in an Alzheimer's animal model without task uncertainty
Humans and macaque monkeys make extra, predictable errors when they switch between visual tasks under uncertainty, a University of Chicago team reports in the 2026 issue of Nature Neuroscience. The researchers, led by co-first author Cheng Xue, link those errors to what they call "feature interference" — a mechanism in which irrelevant visual information leaks into decisions the brain should be ignoring. The paper, DOI 10.1038/s41593-026-02430-w, also points toward a way to fingerprint disorders that erode mental flexibility.
Why does switching tasks produce so many errors?
A half-century of psychology research documents the so-called task-switch cost: people grow slower and make more errors right after rules change. Xue, who trained as a physicist before moving to neuroscience, treated the brain as a dynamical system. "I tend to think of the brain as a dynamical system whose activity is shaped by an underlying landscape, like water flowing over the bedrock of a creek," he told Medical Xpress.
To map that landscape, the team first trained a machine-learning network to reproduce every choice — mistakes included — that monkeys made. They then trained a second network to perform the task correctly. Comparing the two revealed which signals drive errors.
What did the comparison reveal?
The error analysis pointed to a specific mechanism: task-irrelevant visual information bleeds into task-relevant signals when uncertainty is high. "When we are unsure which rule applies, the brain holds on to information it should be ignoring," Xue explained.
The team confirmed this in four ways:
- Under uncertainty, humans remembered the supposedly "irrelevant" feature better.
- The value of the ignored feature pushed choices in a consistent direction.
- In V1, the first stage of visual cortex processing, one feature could be read out from neural signals that normally carry the other.
- Electrically stimulating a single V1 site shifted both judgments together in the direction the model predicted.
What does the visual cortex show under these conditions?
In V1 neurons, the population code for different visual features sits at an angle to one another rather than in perfectly separate channels. Under rule uncertainty, lingering activity from the irrelevant feature tilts the population response, biasing decisions in a predictable direction.
Could error patterns help diagnose brain disorders?
Cognitive flexibility declines with age and in Alzheimer's, Parkinson's, schizophrenia, and substance use disorders. The team believes the new mechanism could eventually support earlier diagnosis and more targeted interventions. "If each condition leaves its own signature in the pattern of errors, those patterns could eventually support earlier diagnosis and more targeted interventions," Xue said.
A related paper in PNAS (2026) by Douglas Ruff, Drew Sheets, and colleagues already reported similar interference in an animal model of Alzheimer's disease, even without task uncertainty. That parallel finding suggests the interference pattern shows up across conditions, not only during rule switches.
What practical advice did the researchers offer?
Xue also offered reassurance for anyone who struggles with multitasking. "It isn't a failure of willpower or discipline," he said. "The brain is prone to relying on the wrong information when switching gears."
His suggestion: finish one task, then take a brief buffer doing something undemanding, like tidying a desk, before starting the next.
The team plans further work testing whether error patterns can act as fingerprints for specific disorders.
via Medical Xpress (Source)
More from Marcus Bennett
Nearby plates
- Brain Cells Vote: How Conflicting Visual Signals Fade in a Fraction of a Second
- MIT Points to Brain Circuit for Updating Beliefs in Schizophrenia
- Fruit Fly Brains Reveal How a Neural Gate Switches Memory On
- One Brain Cell Type Controls How We Learn Movement Timing
- Brain 'Prediction' Signals Are Clocks, Not Anticipation, Mouse Study Finds