Oriel — AI & technology worth noticing, every day.

MAGE: Towards Generalizable Multi-timescale EEG Representations

Paper · https://arxiv.org/abs/2611.09928

arXiv:2611.09928v1 Announce Type: new Abstract: Electroencephalography (EEG) data plays an important role in understanding human brain activity, with various major applications. While current approaches, which largely build on EEG foundation models, have demonstrated strong performance across datasets and tasks, they typically rely on a fixed temporal scale. This design choice limits their ability to capture characteristics of neural dynamic behavior, often spanning across multiple timescales, which are naturally captured in EEG recordings. To address this limitation, we propose MAGE (\textbf{Multi-timescale Adaptive Generalization for EEG}), a foundation model that dynamically adapts to capture and integrate information across different timescales. Specifically, MAGE combines two key mechanisms: (i) a novel multi-timescale attention mechanism, capable of learning representations across multiple temporal scales and (ii) a trainable aggregation module, that dynamically identifies and adapts to the most informative timescale of the input signal. Experimental results across seven datasets demonstrate that MAGE consistently outperforms existing EEG foundation models while using substantially fewer parameters. Moreover, our results support the hypothesis that different downstream tasks benefit from distinct temporal scales. All source code will be released.

First seen by Oriel on .

AI summary

MAGE is an EEG foundation model that adapts its representations across temporal scales, outperforming existing EEG foundation models on seven datasets with substantially fewer parameters.

  • Its multi-timescale attention mechanism learns representations across multiple temporal scales.
  • A trainable aggregation module selects and adapts to the most informative timescale for each input signal.
  • Results support the hypothesis that different downstream tasks benefit from distinct temporal scales.
  • The authors say all source code will be released.

#eeg #foundation models #attention

Generated by gpt-5.6-sol. Check the original for details.

Aliases

SourceExternal IDURL
arXiv AI Papersoai:arXiv.org:2611.09928v1https://arxiv.org/abs/2611.09928 Read discussion