The first dedicated foundation model for electrodermal activity (EDA) data, released alongside the largest public dataset for physiological signal modeling.
March 19, 2026
Original Paper
A foundation model for electrodermal activity data
arXiv · 2603.16878
AI-generated illustration
The Takeaway
By releasing 25,000 hours of curated physiological data and the UME foundation model, this work democratizes research in wearable health sensing. It demonstrates that a domain-specific model can outperform generalist time-series models while using 20x fewer computational resources.
From the abstract
Foundation models have recently extended beyond natural language and vision to timeseries domains, including physiological signals. However, progress in electrodermal activity (EDA) modeling is hindered by the absence of large-scale, curated, and openly accessible datasets. EDA reflects sympathetic nervous system activity and is widely used to infer cognitive load, stress, and engagement. Yet very few wearable devices provide continuous, unobtrusive sensing, and the only large-scale archive to d