Metraipolis › archive › dc_6dca8771bddd

What masking geometry works best for EEG foundation models? — arXiv:2609.33487

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.33487
License (as archived)
CC0-1.0
Archived
2026-10-07T22:37:29.112Z
SHA-256 of the archived text
dfd7edbc4f8a2c9144b6ce22c44012fd6c800cb23e81c75f3304e82ba3630a06
Stable URL of this page
https://metraipolis.com/documento/dc_6dca8771bddd

How to cite

What masking geometry works best for EEG foundation models? — arXiv:2609.33487 — arXiv (en), https://arxiv.org/abs/2609.33487 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6dca8771bddd (Metraipolis archive dc_6dca8771bddd, sha256 dfd7edbc4f8a2c9144b6ce22c44012fd6c800cb23e81c75f3304e82ba3630a06)
BibTeX
@misc{dc6dca8771bddd,
  title = {What masking geometry works best for EEG foundation models? — arXiv:2609.33487},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.33487},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6dca8771bddd; sha256 dfd7edbc4f8a2c9144b6ce22c44012fd6c800cb23e81c75f3304e82ba3630a06},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.33487. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

EEG foundation models hold promise for scalable brain-signal decoding across clinical and cognitive neuroscience applications, yet their pre-training pipelines remain poorly understood. Among design choices, the masking strategy is particularly critical: it determines what the network must predict and from which context. Yet it has never been ablated in isolation, as each new model bundles a new masking strategy with a new backbone and objective. In this paper, we formalize the design choices for spatio-temporal masking strategies and train various models with a single pipeline under varying masking configurations across two SSL frameworks (MAE and JEPA). We then systematically evaluate the resulting 58 pre-trained models on the 12 datasets of OpenEEGBench under a linear probe. Both frameworks agree on an optimal masking configuration and on shared failure modes. Outside these, performance is robust: 11 MAE and 9 JEPA configurations are statistically indistinguishable from the best. We further identify a novel JEPA-specific failure mode, tagged bias-inflation collapse, invisible to standard detectors. With a well-chosen mask, our pipeline reaches REVE-level downstream performance at a fraction of REVE's pre-training compute.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.33487 — 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 2026-10-07T22:14:27.957Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).