Aniruddha Seal, Matthew Hennefarth, Professors Laura Gagliardi and Andrew Ferguson, and Professor Michele Parrinello (Italian Institute of Technology, Genoa) and his group have developed a workflow to train machine-learned potentials beyond Kohn-Sham DFT. At its core is WASP – the Weighted Active Space Protocol — an algorithm that assigns consistent active spaces across diverse geometries, enabling multireference-quality machine learned potentials (MLP) and overcoming a long-standing barrier to incorporating multireference electronic structure in MLPs.
Discover more about this breakthrough and related research here.
