Saved in:
Bibliographic Details
Main Authors: Valenti, Agnes, Park, Ina, Georges, Antoine, Millis, Andrew J., Parcollet, Olivier
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.15741
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910067331170304
author Valenti, Agnes
Park, Ina
Georges, Antoine
Millis, Andrew J.
Parcollet, Olivier
author_facet Valenti, Agnes
Park, Ina
Georges, Antoine
Millis, Andrew J.
Parcollet, Olivier
contents Quantum impurity solvers are the computational bottleneck of quantum embedding approaches to correlated materials, such as dynamical mean-field theory (DMFT). We show that neural networks trained on synthetic, material-agnostic data learn the impurity mapping from hybridization functions and local interactions to Green's functions with quantitative accuracy for both model systems and real materials, providing fast solvers for single- and multi-orbital models. Benchmarks against numerically controlled quantum Monte Carlo show that the method reproduces the Mott transition, multi-orbital phase diagrams of Hubbard-Kanamori models, and the electronic properties of SrVO$_3$ and SrMnO$_3$. The learned solvers achieve orders-of-magnitude speedup and can initialize controlled calculations, dramatically accelerating DMFT while preserving accuracy.
format Preprint
id arxiv_https___arxiv_org_abs_2603_15741
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Neural-Network Quantum Embedding Solvers for Correlated Materials
Valenti, Agnes
Park, Ina
Georges, Antoine
Millis, Andrew J.
Parcollet, Olivier
Strongly Correlated Electrons
Quantum impurity solvers are the computational bottleneck of quantum embedding approaches to correlated materials, such as dynamical mean-field theory (DMFT). We show that neural networks trained on synthetic, material-agnostic data learn the impurity mapping from hybridization functions and local interactions to Green's functions with quantitative accuracy for both model systems and real materials, providing fast solvers for single- and multi-orbital models. Benchmarks against numerically controlled quantum Monte Carlo show that the method reproduces the Mott transition, multi-orbital phase diagrams of Hubbard-Kanamori models, and the electronic properties of SrVO$_3$ and SrMnO$_3$. The learned solvers achieve orders-of-magnitude speedup and can initialize controlled calculations, dramatically accelerating DMFT while preserving accuracy.
title Neural-Network Quantum Embedding Solvers for Correlated Materials
topic Strongly Correlated Electrons
url https://arxiv.org/abs/2603.15741