Outlive
LongevityResearchHub
Disputed — 24/100 Other bioRxiv

How Your Brain Uses Predictions to Shape What You See

A recurrent neuronal model for the effect of predictions on sensory processes

biorxiv (preprint) February 13, 2026 Urgen, B. M.; Basturk, H. I.; Boyaci, H.
TL;DR

This paper presents a computational model of how the brain uses predictions and prior knowledge to filter and interpret sensory information. While the model shows promise in explaining existing behavioral and brain imaging data, it's a theoretical study with no direct longevity relevance and hasn't been peer-reviewed yet.

Credibility Assessment Disputed — 24/100
Study Design
Rigor of the research methodology
4/20
Sample Size
Whether the study was sufficiently powered
2/20
Peer Review
Review status and journal reputation
3/20
Replication
Has this finding been independently reproduced?
6/20
Transparency
Funding disclosure and data availability
9/20
Overall
Sum of all five dimensions
24/100

What this means

This is a theoretical neuroscience paper with no direct bearing on aging or longevity. While the computational model is interesting for understanding how brains use predictions, the lack of peer review and absence of any aging-related mechanism means it's not relevant for longevity researchers.

Red Flags: Not peer-reviewed (preprint only). No original data collected—entirely reanalysis of published studies. Zero citations, suggesting very recent/unknown work. No declared funding or conflicts of interest information provided. No relevance to longevity research. Model validation relies entirely on fitting to existing datasets; true predictive power unknown.

This preprint addresses a fundamental neuroscience question: how does the brain use predictions and expectations to shape sensory perception? The authors propose a recurrent neural network model—a simplified mathematical representation of how neurons might interact—to explain this process. The model assumes that predictions flow downward to sensory regions while actual sensory signals flow upward, and that these signals are integrated through feedback loops in established brain circuits.

The researchers tested their model against published behavioral and fMRI data from three different studies involving visual recognition tasks (faces, houses, and other stimuli). They fit the model to one dataset, then checked whether the same optimized parameters could predict findings from other studies. The model successfully reproduced the main findings across diverse stimulus types, suggesting it captures something generalizable about how prediction influences perception.

However, this work is purely theoretical and computational—it uses existing published data, not new experiments. The authors did not conduct original studies, collect new brain imaging data, or test the model's predictions experimentally. The paper is also a preprint, meaning it has not undergone peer review at a traditional journal. While the approach is scientifically sound, peer review by independent experts would strengthen confidence in the claims.

Critically, this research has no connection to longevity science. Predictive processing in sensory perception is interesting for cognitive neuroscience and AI, but doesn't address aging, lifespan, healthspan, or any mechanisms of aging. The paper would not inform development of longevity interventions or biomarkers.

The model's generalizability across studies is a modest strength, but the lack of novel empirical validation, absence of peer review, and complete disconnect from aging biology make this unsuitable for longevity researchers. The work may eventually inform digital brain models or computational understanding of aging-related cognitive changes, but that application is speculative and distant.

Digital Twins Machine Learning
View Original Source

0 Comments