The abstract is read at the publisher; the summary is JClub's.
Precision-weighted predictions and information-theoretic surprisal dynamically modulate human auditory cortex frequency tuning, with sharpening and multiplicative gain changes explaining significantly more variance than static pRF baselines across primary, belt, and associative areas.
- Why it matters: Standard decoding mainly reflects overall signal-to-noise, leaving tuning-width versus response-gain contributions unclear for predictive perception in the auditory hierarchy.
- What they did: Using 7T fMRI, population receptive field modeling, and a continuous Bayesian ideal observer, the authors compared static pRF tuning to two tone-by-tone dynamic models: precision-driven sharpening of tuning width and surprisal-driven gain scaling.
- The result: Both dynamic models explained significantly more variance than static baselines, counterfactual inversions added no improvement, and local precision sharpened expected-frequency pRFs while surprisal acted as a global broadcast multiplicatively scaling pRF gain across tonotopic channels.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.