Background
Electroencephalography (EEG) foundation models (FMs) are trained to reconstruct or contrastively align short patches, then pooled into a fixed embedding. This study tests whether these embeddings retain the long-range temporal correlations (LRTC) quantified by the detrended-fluctuation-analysis (DFA) exponent and whether it governs cross-population transfer.
Methodology
We probed five EEG FMs spanning raw-waveform and spectral-input architectures (REVE, LaBraM, BENDR, CBraMod, BIOT) on two out-of-distribution cohorts, comparing recovery of the DFA exponent against the static 1/f aperiodic slope. Order-preserving and residualization controls tested for pooling or aperiodic shadowing. A montage-harmonized, zero-shot transfer task compared the frozen embedding with the DFA exponent across three cohorts (adding a Western reference).
Main Findings
None of the five FMs represented the LRTC in the temporal order. Raw-waveform models (REVE, LaBraM, BENDR) recovered neither the DFA exponent nor the 1/f slope (R^2 < 0.01).
Blogger's Review: This study highlights the shortcomings of EEG foundation models in capturing long-range temporal correlations, emphasizing the fragility of model design in cross-population applications. These findings provide crucial directions for future improvements in EEG models, particularly in capturing and utilizing long-range temporal information.