Metraipolis › archive › dc_cd0583f5e869
NeurDuo-EEG: A Long-Sequence EEG Foundation Model with Persistent State and Explicit Memory — arXiv:2609.38587
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.38587
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T13:45:23.346Z
- SHA-256 of the archived text
3449936b82010579a28d2a828510461546e358a6ff5d05cc8cfed37219ac02fe- Stable URL of this page
https://metraipolis.com/documento/dc_cd0583f5e869
How to cite
NeurDuo-EEG: A Long-Sequence EEG Foundation Model with Persistent State and Explicit Memory — arXiv:2609.38587 — arXiv (en), https://arxiv.org/abs/2609.38587 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_cd0583f5e869 (Metraipolis archive dc_cd0583f5e869, sha256 3449936b82010579a28d2a828510461546e358a6ff5d05cc8cfed37219ac02fe)
BibTeX
@misc{dccd0583f5e869,
title = {NeurDuo-EEG: A Long-Sequence EEG Foundation Model with Persistent State and Explicit Memory — arXiv:2609.38587},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.38587},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_cd0583f5e869; sha256 3449936b82010579a28d2a828510461546e358a6ff5d05cc8cfed37219ac02fe},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.38587. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Electroencephalography (EEG) is recorded continuously over hours, with relevant dynamics spanning timescales from milliseconds to hours. Most EEG foundation models nevertheless process fixed windows independently, limiting their ability to capture information encoded in long-timescale dynamics. State-space architectures enable persistent recurrent processing, but long-range information remains implicitly compressed in recurrent states. We present NeurDuo-EEG, a causal EEG foundation model with channel-resolved persistent memory. NeurDuo-EEG introduces multi-timescale memory management with learned consolidation and selective retrieval, enabling persistent modelling of continuous EEG with fixed-size state. It is pre-trained on 3,955 hours of EEG from 17 public datasets using multichannel autoregressive prediction of discrete spectral codes. Across three short-window and two long-sequence downstream tasks, NeurDuo-EEG achieves the best performance on four of five benchmarks, including all three short-window tasks and seizure detection, where AUC-PR improves from $0.285$ to $0.471$ over the strongest non-NeurDuo baseline. NeurDuo-EEG also remains competitive on sleep staging and supports efficient streaming inference, with nearly constant per-chunk latency as the available history grows to one hour. Notably, the Small variant achieves this with only 4.7M backbone parameters. These results demonstrate the value of persistent, multi-timescale modelling for both long-sequence and short-window EEG analysis. Our code is available at this https URL . Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.LG). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.38587 — 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 2026-10-07T13:18:10.226Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).