The abstract is read at the publisher; the summary is JClub's.
In six adults watching Friends, participant-specific fMRI hidden Markov models identified 42–47 recurring brain states across 38–57 hours, aligned to episode moments.
- Why it matters: It tests whether a brain’s repertoire and input-dependent expression persist over long, naturalistic experience, and whether state identity generalizes across content and sensory formats.
- What they did: They fit hidden Markov models to separate portions of viewing and used temporal screening, then related state identity to features from pretrained video, audio, and language networks.
- The result: States matched across participants tended to recur at the same episode moments, and intermediate-to-late layer features predicted state identity; models generalized to new films and language predictions transferred to reading and audiobook listening.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.