A 12.9M-parameter audio-to-fMRI encoder achieves zero-shot prediction of speech-evoked brain responses on 324 unseen participants and few-shot adaptation with ~10 minutes of data, outperforming larger baselines and per-participant ridge regression.
Transformer brain encoders explain human high-level visual responses
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
RABBiT: Rapidly adaptive BOLD foundation model via brain-tuning for accurate zero-shot and few-shot prediction of speech-elicited responses in the brain
A 12.9M-parameter audio-to-fMRI encoder achieves zero-shot prediction of speech-evoked brain responses on 324 unseen participants and few-shot adaptation with ~10 minutes of data, outperforming larger baselines and per-participant ridge regression.