EEG Foundation Models: Progresses, Benchmarking, and Open Problems

TL;DR

This paper reviews and benchmarks EEG foundation models, analyzing their design choices, evaluating their performance across diverse datasets and paradigms, and identifying open challenges in transferability and model scaling.

Contribution

It provides a comprehensive taxonomy of EEG foundation models, evaluates 12 models across multiple tasks, and highlights key insights on transferability, model size, and training strategies.

Findings

Linear probing often insufficient for transfer tasks.

Specialist models trained from scratch remain competitive.

Larger models do not always improve performance.

Abstract

Electroencephalography (EEG) foundation models have recently emerged as a promising paradigm for brain-computer interfaces (BCIs), aiming to learn transferable neural representations from large-scale heterogeneous recordings. Despite rapid progresses, there lacks fair and comprehensive comparisons of existing EEG foundation models, due to inconsistent pre-training objectives, preprocessing choices, and downstream evaluation protocols. This paper fills this gap. We first review 50 representative models and organize their design choices into a unified taxonomic framework including data standardization, model architectures, and self-supervised pre-training strategies. We then evaluate 12 open-source foundation models and competitive specialist baselines across 13 EEG datasets spanning nine BCI paradigms. Emphasizing real-world deployments, we consider both cross-subject generalization under a leave-one-subject-out protocol and rapid calibration under a within-subject few-shot setting. We further compare full-parameter fine-tuning with linear probing to assess the transferability of pre-trained representations, and examine the relationship between model scale and downstream performance. Our results indicate that: 1) linear probing is frequently insufficient; 2) specialist models trained from scratch remain competitive across many tasks; and, 3) larger foundation models do not necessarily yield better generalization performance under current data regimes and training practices.