Deep Mind 21 functional does not extrapolate to transition metal chemistry
Heng Zhao, Tim Gould, Stefan Vuckovic
Abstract
, 1385-1389], trained on main-group chemistry, extrapolates to transition metal chemistry (TMC). We show that DM21 demonstrates comparable or occasionally superior accuracy to B3LYP for TMC, but consistently struggles with achieving self-consistent field convergence for TMC molecules. We also compare main-group and TMC machine-learning DM21 features to shed light on DM21's challenges in TMC. We finally propose strategies to overcome limitations in the extrapolative capabilities of machine-learned functionals in TMC.
Topics & Concepts
ExtrapolationDensity functional theoryChemistryArtificial intelligenceComputational chemistryComputer scienceMathematicsStatisticsMachine Learning in Materials ScienceComputational Drug Discovery MethodsProtein Structure and Dynamics