Yi Cao’s paper, “Migration as a Probe: A Generalizable Benchmark Framework for Specialist vs. Generalist Machine-Learned Force Fields,” has been selected for a Spotlight Presentation at the AI4Mat Workshop, NeurIPS 2025 (top ~10% of accepted papers).
The work introduces a new paradigm for evaluating ML force fields—combining thermodynamic and kinetic benchmarks to probe how models truly learn and generalize. By using migration trajectories as dynamic tests, the study moves beyond static accuracy metrics and provides deeper insight into model reliability and physical interpretability.
Congratulations to Yi for this recognition at one of the world’s leading AI conferences.
Paper Link: https://arxiv.org/abs/2509.00090
Code available at: https://github.com/yicao-elina/MigrationBench