Fig. 1: Main features of SpookyNet and problems addressed in this work.
From: SpookyNet: Learning force fields with electronic degrees of freedom and nonlocal effects
a Optimized geometries of Ag\({}_{3}^{+}\)/Ag\({}_{3}^{-}\) (left) and singlet/triplet CH2 (right). Without information about the electronic state (charge/spin), machine learning models are unable to distinguish between the different structures. b Au2 dimer on a MgO(001) surface doped with Al atoms (Au: yellow, Mg: gray, O: red, Al: pink). The presence of Al atoms in the crystal influences the electronic structure and affects Au2 binding to the surface, an effect which cannot be adequately described by only local interactions. c Potential energy Epot (solid black) for O–H bond dissociation in water. The asymptotic behavior of Epot for very small and very large bond lengths can be well-approximated by analytical short-ranged Esr (dotted red) and long-ranged Elr (dotted orange) energy contributions, which follow known physical laws. When they are subtracted from Epot, the remaining energy (solid blue) covers a smaller range of values and decays to zero quicker, which simplifies the learning problem. d Visualization of a random selection of learned interaction functions for SpookyNet trained on the QM7-X71 dataset. They are designed to closely resemble atomic orbitals, facilitating SpookyNet’s ability to extract chemical insight from data.
