Active Learning Paper Accepted at the NeurIPS 2026 AI4Mat Workshop

Our paper “Accelerating Materials Discovery with Active Thermodynamic Surface Shape Mapping” has been accepted at the AI for Accelerated Materials Design (AI4Mat) Workshop at NeurIPS 2026. The work, a collaboration with Andrew Tipping, Bobby Huggins, and Roman Garnett that includes Ray Yang, introduces an active learning acquisition function that efficiently locates features such as phase transitions on thermodynamic surfaces, with a test on phase diagram learning.