Virtual orbital many-body expansions: A possible route towards the full configuration interaction limit
Janus J. Eriksen, Filippo Lipparini, J\"urgen Gauss

TL;DR
This paper introduces a virtual orbital many-body expansion method that approximates full configuration interaction energies with high accuracy, using reduced orbital spaces and parallel computation, demonstrated on water molecules.
Contribution
The paper presents a novel approach to approximate FCI energies efficiently by decomposing in virtual orbitals and employing a screening protocol, enabling calculations in reduced orbital subspaces.
Findings
Achieved sub-kJ/mol accuracy for FCI energies in water.
Demonstrated scalability with basis set size from double- to quadruple-zeta.
Implemented embarrassingly parallel computation for energy calculations.
Abstract
In the present letter, it is demonstrated how full configuration interaction (FCI) results in extended basis sets may be obtained to within sub-kJ/mol accuracy by decomposing the energy in terms of many-body expansions in the virtual orbitals of the molecular system at hand. This extension of the FCI application range lends itself to two unique features of the current approach, namely that the total energy calculation can be performed entirely within considerably reduced orbital subspaces and may be so by means of embarrassingly parallel programming. Facilitated by a rigorous and methodical screening protocol and further aided by expansion points different from the Hartree-Fock solution, all-electron numerical results are reported for HO in polarized core-valence basis sets ranging from double- (10 , 28 ) to quadruple- (10 , 144 ) quality.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
