Channel Adaptation for EEG Foundation Models: A Systematic Benchmark Across Architectures, Tasks, and Training Regimes
Kuntal Kokate, Bruno Aristimunha, Dung Truong, Arnaud Delorme · 2026
Read Channel Adaptation for EEG Foundation Models: A Systematic Benchmark Across Architectures, Tasks, and Training Regimes by Kuntal Kokate, Bruno Aristimunha, Dung Truong, Arnaud Delorme (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-23091.txt.
Description
Kokate and three colleagues systematically compare four channel adaptation methods for EEG foundation models across five pretrained models, five downstream tasks, and two training regimes. Rigid-montage models require external adaptation, while flexible models match or beat it unaided when fine-tuned, yet gain from outside help under frozen encoders. External adaptation can cause severe negative transfer when fine-tuning flexible models. The best method depends on architecture, and a compact 5-million-parameter model beats models up to 31 times larger across four of five data sets.
Catalog record
| CREATOR | Kuntal Kokate, Bruno Aristimunha, Dung Truong, Arnaud Delorme |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 27982 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.23091