Dataset for "Automated discovery of algorithms for molecular electronic structure calculations using physics-informed program synthesis"

Read full text Summarise with VIRAT
Authors: Kyle Acheson, Rastislav Turányi, Scott Habershon
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

We demonstrate a physics-informed program synthesis (PIPS) approach that can be used to identify entirely-new algorithms that approximate the results of single-reference electronic structure approaches like Hartree-Fock (HF) and density-functional theory (DFT) - but without any self-consistent field iterations at all. Our PIPS strategy exploits the fact that the eigenvectors of the Fock matrix F (or Kohn-Sham matrix K) are the same as the eigenvectors of a broad class of matrix functions, f(F). As a result, PIPS can be used to seek matrices M that yield the same molecular orbital coefficients as converged HF or DFT calculations. We demonstrate this approach by generating new algorithms that accurately predict total energies for a series of heterodiatomic molecules (LiCl, LiF, NaCl, NaF) and C1-C4 hydrocarbons; further simulations of C8-C20 alkane species demonstrate further transferability and efficiency of the resulting algorithms. We obtain novel algorithms that can reproduce HF or DFT energies to within 0.1 kcal/mol/atom while requiring only a single matrix-diagonalization operation, rather than an iterative self-consistent field convergence. The approach demonstrated here could be similarly applied to more complex wavefunction ansatze, opening an interesting optimization-based pathway to identifying accurate yet efficient algorithms for molecular quantum chemistry. Supporting data for "Automated discovery of algorithms for molecular electronic structure calculations using physics-informed program synthesis”, including calculation data for diatomic molecules and alkane molecules. Both raw (unprocessed) and final (compiled) datasets are included.

Fields of Study & Subjects

Advanced Chemical Physics Studies Machine Learning in Materials Science Computational Drug Discovery Methods

Publication details

Journal / source
Warwick Research Archive Portal (University of Warwick)
Publication date
4 March 2026
Language
EN
Document type
Dataset