AI Paper Atlas

Research map / Video / Video segmentation and tracking

MAGE: Towards Generalizable Multi-timescale EEG Representations

Tuan-Kiet Doan, Van Chien Nguyen, Trung-Hieu Tran, Quang Hung Pham, Ngoc-Son Vu, Huy Phan, Phi Le Nguyen, Duc-Han LE, Nikola Simidjievski, Van-Tam Nguyen

NeurIPS 2026 · Video › Video segmentation and tracking · approach: Sign language and physiological video analysis · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes MAGE, an EEG representation approach that captures multi-timescale neural dynamics instead of relying on a fixed temporal scale in EEG foundation models.

Problem

Current EEG foundation models rely on a fixed temporal scale, which limits their ability to capture neural dynamics that span multiple timescales.

Approach

Moves beyond fixed-scale EEG encoding to learn multi-timescale representations intended to generalize across datasets and tasks (details beyond the abstract not available).

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use multi-timescale tokens with adaptive aggregator instead of fixed-patch foundation modelsMAGE · BCIC-IV-2a, PhysioNet-MI, FACED, TUAB, KaggleERNbalanced accuracy: e.g. PhysioNet-MI 59.96 -> 62.31 vs CSBrainImproved
use multi-timescale model on Sleep-EDFx and SHU-MIMAGE · Sleep-EDFx, SHU-MIbalanced accuracy: Sleep-EDFx 70.39 (CBraMod) vs 69.98; SHU-MI 63.22 (LaBram) vs 62.31No clear effect
MAGE with fewer parametersMAGE · seven EEG datasetsaverage rank: table: 1.81 vs 2.71 (CSBrain); text states 1.43 vs 2.81Mixed
combine all three timescales instead of a single timescale (ablation)MAGE · FACEDbalanced accuracy: 46.65 (1s only) -> 51.30Improved
combine all three timescales instead of a single timescale (ablation)MAGE · Sleep-EDFxbalanced accuracy: 67.99 (1s only) -> 70.24Improved
route to only the top-weighted timescale instead of weighted sum (ablation)MAGE-router · FACED, PhysioNet-MI, Sleep-EDFx, SHU-MIbalanced accuracy: FACED 51.30 -> 49.87; PhysioNet-MI 61.15 -> 58.83Hurt
train aggregator downstream instead of freezing (ablation)MAGE · Sleep-EDFxbalanced accuracy: 66.16 -> 70.24Improved
train aggregator downstream instead of freezing (ablation)MAGE · PhysioNet-MIbalanced accuracy: 61.61 (frozen) vs 61.15 (trained)Hurt
add Gaussian noise to aggregator logits during pretraining (ablation)MAGE · Sleep-EDFxbalanced accuracy: 68.79 -> 70.24Improved
add Gaussian noise to aggregator logits during pretraining (ablation)MAGE · PhysioNet-MIbalanced accuracy: 61.57 -> 61.15Hurt
learned aggregator instead of uniform weights (ablation)MAGE · Sleep-EDFxbalanced accuracy: 70.00 -> 70.24Improved
aggregator instead of Medformer inter attention (ablation)MAGE · SHU-MIbalanced accuracy: 63.52 -> 62.43Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy