Reproduce “ZeroMAG: Zero-Shot Multimodal Adapter Generation for Plug-and-Play 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
Yubo Wang, Jingying Ma, Xinliang Zhou, Yangxuan Zhou, Jiquan Wang, Sha Zhao, Yiyuan Yang, Yi Ding and 3 more
ZeroMAG generates recording-conditioned adapters to add companion physiological signals to a frozen EEG model. It uses unlabeled calibration recordings and task context, with functional supervision. Across six held-out datasets and three separately trained backbones, balanced accuracy improves by 7.22 points over EEG-only inference and 4.89 over direct weight regression. “Zero-shot” excludes target labels and optimization, but uses target recordings.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.