Search papers, labs, and topics across Lattice.
This paper presents a Hamiltonian learning framework leveraging superposition-of-atomic-potentials (SAP) features to enhance the machine learning of electronic Hamiltonians, which is crucial for predicting electronic wave functions and physical observables. By utilizing a symmetry-adapted intrinsic atomic orbital learning basis and an orbital-based graph neural network, the model effectively predicts converged Kohn-Sham Fock matrices and extends to larger basis sets through a downfolding scheme. The approach demonstrates high accuracy on the QM9 dataset and for organic charge-transport materials, achieving a mean absolute error of 4.8 meV for unseen substituted-benzene heterodimers, highlighting its potential for scalable electronic-structure predictions.
SAP-based machine learning achieves unprecedented accuracy in predicting electronic structures, even for unseen molecular configurations.
Machine learning (ML) of electronic Hamiltonians offers a unified route to electronic wave functions and physical observables. We introduce a Hamiltonian learning framework built on electronic features derived from the superposition-of-atomic-potentials (SAP) approximation, an efficient self-consistent-field initial guess that captures essential electron-electron screening. SAP quantities define a symmetry-adapted intrinsic atomic orbital learning basis and provide physics-informed inputs to an orbital-based graph neural network that predicts converged Kohn-Sham Fock matrices. To extend the approach to larger basis sets, we further develop a downfolding scheme that predicts large-basis electronic structure from minimal-basis features. On the QM9 dataset, the model accurately reproduces frontier and core orbital energies, dipole moments, and the full density of states. For organic charge-transport materials, it yields accurate intermolecular transfer integrals for benzene, tetracyanoquinodimethane (TCNQ), and tetrathiafulvalene (TTF) dimers, and transfers to unseen substituted-benzene heterodimers with a mean absolute error of 4.8 meV. These results establish SAP-based ML of electronic Hamiltonians as a transferable and scalable tool for high-throughput electronic-structure prediction.
