AI RESEARCH
Beyond Accuracy: Robustness, Interpretability and Expressiveness of EEG Foundation Models
arXiv CS.LG
•
ArXi:2605.17562v1 Announce Type: new EEG foundation models (EEG-FMs) have been evaluated predominantly on clean, in-distribution accuracy, leaving their robustness, interpretability and representational quality largely unexamined. This study addresses these gaps by benchmarking six EEG-FMs against a baseline deep learning model across eight datasets. Beyond clean accuracy, we conduct three layers of analysis: (i) Robustness: we apply test-time perturbations including additive noise, random and region-based channel dropout and region-specific noise injection.