Reproduce “Neural State Prediction: Obstructing Shortcut Learning in EEG Foundation Models” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Kieren Yu, Ziyang Liu, Chang Huang, Jintai Chen, Kaishun Wu
Neural State Prediction (NSP) constrains self-supervised EEG prediction by residualizing additive channel and relative-time effects in target representations and masking nearby channel and time content from context. Under EEG-FM-Bench multi-task full-parameter fine-tuning, it achieves 63.94% macro balanced accuracy, 2.35 percentage points above the strongest evaluated baseline. This result is specific to that protocol; performance varies across tasks and transfer settings.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.