Spatial Signatures\nof Electron Correlation in Least-Squares\nTensor Hypercontraction

Least-squares tensor hypercontraction\n(LS-THC) has received\nsome\nattention in recent years as an approach to reduce the significant\ncomputational costs of wave function-based methods in quantum chemistry.\nHowever, previous work has demonstrated that LS-THC factorization\nperforms disproportionately worse in the description of wave function\ncomponents (e.g., cluster amplitudes <i>T̂</i><sub>2</sub>) than Hamiltonian components (e.g., electron repulsion integrals\n(<i>pq</i>|<i>rs</i>)). This work develops novel\ntheoretical methods to study the source of these errors in the context\nof the real-space <i>T̂</i><sub>2</sub> kernel, and\nreports, for the first time, the existence of a “correlation\nfeature” in the errors of the LS-THC representation of the\n“exchange-like” correlation energy <i>E</i><sub><i>X</i></sub> and <i>T̂</i><sub>2</sub> that is remarkably consistent across ten molecular species, three\ncorrelated wave functions, and four basis sets. This correlation feature\nportends the existence of a “pair point kernel” missing\nin the usual LS-THC representation of the wave function, which critically\ndepends upon pairs of grid points situated close to atoms and with\ninterpair distances between one and two Bohr radii. These findings\npoint the way for future LS-THC developments to address these shortcomings.

Publication Details

Journal
Figshare
Published
2025-01-09
DOI
https://doi.org/10.1021/acs.jpca.4c06666.s001
Primary Topic
Tensor decomposition and applications
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Spatial Signatures\nof Electron Correlation in Least-Squares\nTensor Hypercontraction

Figshare
Tensor decomposition and applications
article

Spatial Signatures\nof Electron Correlation in Least-Squares\nTensor Hypercontraction

article en

Abstract

Least-squares tensor hypercontraction\n(LS-THC) has received\nsome\nattention in recent years as an approach to reduce the significant\ncomputational costs of wave function-based methods in quantum chemistry.\nHowever, previous work has demonstrated that LS-THC factorization\nperforms disproportionately worse in the description of wave function\ncomponents (e.g., cluster amplitudes <i>T̂</i><sub>2</sub>) than Hamiltonian components (e.g., electron repulsion integrals\n(<i>pq</i>|<i>rs</i>)). This work develops novel\ntheoretical methods to study the source of these errors in the context\nof the real-space <i>T̂</i><sub>2</sub> kernel, and\nreports, for the first time, the existence of a “correlation\nfeature” in the errors of the LS-THC representation of the\n“exchange-like” correlation energy <i>E</i><sub><i>X</i></sub> and <i>T̂</i><sub>2</sub> that is remarkably consistent across ten molecular species, three\ncorrelated wave functions, and four basis sets. This correlation feature\nportends the existence of a “pair point kernel” missing\nin the usual LS-THC representation of the wave function, which critically\ndepends upon pairs of grid points situated close to atoms and with\ninterpair distances between one and two Bohr radii. These findings\npoint the way for future LS-THC developments to address these shortcomings.

Figshare
Openalex Percentile: Top 60%
Tensor decomposition and applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Spatial Signatures\nof Electron Correlation in Least-Squares\nTensor Hypercontraction · Figshare (2025) | TGRS Research Map | TGRS