Sample what you need: efficient active-learning strategies for accurate machine-learned potentials
Date:
Invited talk at Multi-Scale Condensed Phase Simulations: Methods and Applications, Telluride, CO, USA
Invited talk at the Telluride Science workshop Multi-Scale Condensed Phase Simulations: Methods and Applications (13–17 July 2026), on active-learning strategies that select only the configurations worth computing when training machine-learning interatomic potentials.