Metraipolis › archive › dc_d8de26a8dee4

Matched-Input Estimates Differ in Sign Across Architectures: Auditing EEG Foundation Models on Motor Imagery — arXiv:2609.23924

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.23924
License (as archived)
CC0-1.0
Archived
2026-10-09T07:07:27.200Z
SHA-256 of the archived text
0725adb2f84e5310dcce52c2eb23d7a91494db38124f657e2f548cb3aadcfcb5
Stable URL of this page
https://metraipolis.com/documento/dc_d8de26a8dee4

How to cite

Matched-Input Estimates Differ in Sign Across Architectures: Auditing EEG Foundation Models on Motor Imagery — arXiv:2609.23924 — arXiv (en), https://arxiv.org/abs/2609.23924 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d8de26a8dee4 (Metraipolis archive dc_d8de26a8dee4, sha256 0725adb2f84e5310dcce52c2eb23d7a91494db38124f657e2f548cb3aadcfcb5)
BibTeX
@misc{dcd8de26a8dee4,
  title = {Matched-Input Estimates Differ in Sign Across Architectures: Auditing EEG Foundation Models on Motor Imagery — arXiv:2609.23924},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.23924},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d8de26a8dee4; sha256 0725adb2f84e5310dcce52c2eb23d7a91494db38124f657e2f548cb3aadcfcb5},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.23924. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Pretrained EEG foundation models are increasingly proposed as general-purpose encoders for brain-computer interfaces, yet recent benchmarks disagree about when their representations transfer to downstream tasks. We audit LaBraM and CBraMod on motor imagery under a validation-locked protocol in which preprocessing, architecture, optimization, freeze depth, checkpoint, temperature, and method selection are determined using training-session data only. On four-class BCI Competition IV-2a, every supervised comparator evaluated here outperforms every foundation-model configuration, including validation-selected fine-tuning. We then examine a key confound: foundation models and task-specific decoders are normally evaluated using different input pipelines. Retraining three supervised architectures on the broadband arrays consumed by the foundation models produces matched-input accuracy differences of opposite sign across architectures: broadband input improves ATCNet by 0.078 accuracy while reducing EEG Conformer accuracy by 0.088. None of the three individual matched-input terms is significant after multiple-comparison correction at n = 9, so we treat the sign variation descriptively rather than as a formal architecture-by-pipeline interaction. These observed sign differences suggest that a single comparator may not provide an architecture-invariant decomposition of a pretrained-versus-supervised performance gap. The four-class deficit also does not reproduce uniformly across motor-imagery datasets: on two-class BNCI2014-004 we cannot detect the same separation between fine-tuned CBraMod and the supervised comparators. Finally, validation-fitted temperature scaling returns foundation-model calibration error to the supervised range despite substantially lower four-class accuracy.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-22 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Neurons and Cognition (q-bio.NC).
Pagina del lavoro: https://arxiv.org/abs/2609.23924 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-09T06:58:09.352Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).