Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:1901.01501 · Physical review. B./Physical review. B

Accelerated continuous time quantum Monte Carlo method with machine learning

Taegeun Song, Hunpyo Lee

Published Jul 31, 2019• 12 citations
arXivHTMLPDF

Abstract

An acceleration of continuous time quantum Monte Carlo (CTQMC) methods is a potentially interesting branch of work as they are matchless as impurity solvers of a dynamical mean field theory (DMFT) approach for the description of strongly correlated systems. The inversion of the $k\ifmmode\times\else\texttimes\fi{}k$ matrix with ${k}^{2}$ operations given by the diagram expansion order $k$ in the CTQMC fast update and the multiplication of the $k\ifmmode\times\else\texttimes\fi{}k$ matrix, and the noninteracting properties with $k\ifmmode\times\else\texttimes\fi{}{\ensuremath{\omega}}_{{n}_{\text{max}}}^{m\ensuremath{-}1}$ operations to measure the $m$-point correlators, are computationally time consuming. Here we propose the CTQMC method in combination with a machine learning technique, which eliminates the $k\ifmmode\times\else\texttimes\fi{}{\ensuremath{\omega}}_{{n}_{\text{max}}}$ and $k\ifmmode\times\else\texttimes\fi{}{\ensuremath{\omega}}_{{n}_{\text{max}}}^{3}$ operations for the two-point impurity Green's functions ${G}_{\ensuremath{\sigma}}(i{\ensuremath{\omega}}_{n})$ and four-point vertices ${\ensuremath{\chi}}_{\ensuremath{\sigma},\overline{\ensuremath{\sigma}}}(i{\ensuremath{\omega}}_{n{}_{1}},i{\ensuremath{\omega}}_{n{}_{2}},i{\ensuremath{\omega}}_{n{}_{3}},i{\ensuremath{\omega}}_{n{}_{4}})$, respectively. This method not only predicts the accurate physical properties at low temperature, but also dramatically decreases the computational times of ${\ensuremath{\chi}}_{\ensuremath{\sigma},\overline{\ensuremath{\sigma}}}(i{\ensuremath{\omega}}_{n{}_{1}},i{\ensuremath{\omega}}_{n{}_{2}},i{\ensuremath{\omega}}_{n{}_{3}},i{\ensuremath{\omega}}_{n{}_{4}})$ for the nonlocal extension of DMFT approximation.

DOI
10.1103/physrevb.100.045153

Related papers

Continuous Time Quantum Monte Carlo in Combination with Machine Learning on the Hubbard Model

Hunpyo Lee

The acceleration of exact continuous time quantum Monte Carlo (CTQMC) approaches in multi-site or multi-orbital systems is extremely interesting work, because these approaches are very time-consuming in terms of numerical computation and might account for the nature of exotic behaviors such as high-temperature superconductivity and Mott insulator behavior observed in the strongly correlated materials. We extend the recently developed interaction-expansion CTQMC method in combination with a machine learning (CTQMC+ML) approach for the single-site and single-orbital systems to multi-site and multi-orbital ones. This method can be applied to explore the nonlocal correlation effects in lattice models and to study the electronic structure of real materials via an ab-initio density functional theory plus dynamical mean field theory approach. We find that our CTQMC+ML method for multi-site (and multi-orbital) systems accurately predicts the impurity Green’s function with less computational time than the CTQMC approaches, as in the case of the single-site and single-orbital version.

Journal of the Korean Physical Society•Nov 1, 2019•1 citation

Continuous-time quantum Monte Carlo solver for dynamical mean field theory in the compact Legendre representation

Evan Sheridan, Cédric Weber, Evgeny A. Plekhanov +1 more

Dynamical mean-field theory (DMFT) is one of the most widely-used methods to treat accurately electron correlation effects in ab-initio real material calculations. Many modern large-scale implementations of DMFT in electronic structure codes involve solving a quantum impurity model with a continuous-time quantum Monte Carlo (CT-QMC) solver [Rubtsov et al., Phys. Rev. B 72, 035122 (2005); Werner et al., Phys. Rev. Lett. 97, 076405 (2006); Werner and Millis, Phys. Rev. B 74, 155107 (2006); Gull et al., Rev. Mod. Phys. 83, 349 (2011)]. The main advantage of CT-QMC is that, unlike standard quantum Monte Carlo approaches, it is able to generate the local Green's functions $G(\ensuremath{\tau})$ of the correlated system on an arbitrarily fine imaginary time $\ensuremath{\tau}$ grid, and is free of any systematic errors. In this work, we extend a hybrid QMC solver proposed by Khatami et al. [Phys. Rev. E 81, 056703 (2010)] and Rost et al. [Phys. Rev. E 87, 053305 (2013)] to a multiorbital context. This has the advantage of enabling impurity solver QMC calculations to scale linearly with inverse temperature $\ensuremath{\beta}$, and permit its application to $d$- and $f$-band materials. In addition, we present a Green's-function processing scheme which generates accurate quasicontinuous imaginary time solutions of the impurity problem which overcome errors inherent to standard QMC approaches. This solver and processing scheme are incorporated into a full DFT+DMFT calculation using the castep DFT code [Clark et al., Z. Kristallogr. 220, 567 (2005)]. Benchmark calculations for ${\mathrm{SrVO}}_{3}$ properties are presented. The computational efficiency of this method is also demonstrated.

arXiv:1811.12869•Physical review. B./Physical review. B•May 30, 2019•14 citations

Acceleration of CTQMC via Denoising Diffusion Models with Auxiliary Spectral Functions

Rishi Rao, Li Zhu

Continuous-time Quantum Monte Carlo (CTQMC) impurity solvers are a numerically exact, but expensive, option for solving the Dynamical Mean Field Theory (DMFT) equations, with statistical noise from undersampling presenting a major challenge to low temperature studies. This noise is especially egregious within the sampled self-energy, which can destabilize the self-consistency loop upon insertion into the Dyson equation. In this work, we introduce a neural denoiser for fermionic imaginary time Green's functions based on the diffusion model architecture trained on synthetic data. The network predicts positive and normalized auxiliary spectral functions consistent with the noisy imaginary axis CTQMC data, which is then refined via autodifferentiation to match the measured self-energy, its first two high-frequency moments, and impurity occupation similar to the Constrained Residual Minimization (CRM) method. This leads to Green's functions and self-energies that are causal by design. Through this denoising framework we are able to reduce the required CTQMC measurement steps by a factor of 10-100. We benchmark our method with a half-filled Hubbard model on a Bethe lattice, a two-patch Dynamical Cluster Approximation (DCA) of the two-dimensional Hubbard model, and the SrVO$_3$ system. We reproduce the Mott transition upon increasing Coulomb interaction for the half-filled Hubbard model with only 5000 CTQMC measurements, the pseudogap transition for the DCA two-dimensional Hubbard model with only 5$\times$10$^5$ measurements, and match the spectral function of SrVO$_3$ with 50 times fewer measurements than the reference. Since we generate auxiliary spectral functions, we recover the approximate spectral functions of the systems for free without performing analytic continuation.

arXiv:2610.28542•Oct 20, 2026•0 citations

Continuous-Time Quantum Monte Carlo Impurity Solvers: Improvements and Applications

P. Sémon

Abstract: Originally designed for the study of strong electronic correlations in model Hamiltonians, dynamical mean field theory (DMFT) has become, in combination with density functional theory (DFT), a powerful tool for ab initio simulations of real materials. At the heart of DMFT lies the solution of a quantum impurity problem. While only the continuous-time quantum Monte Carlo (CT-QMC) impurity solvers yield (statistically) exact solutions of a general impurity problem, they are quite complex and computationally expensive. Hence, in this thesis we are interested in improving the CT-QMC impurity solvers. After a short introduction to DMFT and its cluster extensions, we begin by reviewing two of the CT-QMC impurity solvers, the interaction expansion or “Rubtsov” solver (CT- INT) and the hybridization expansion solver (CT-HYB). Focussing on the latter, which is the algorithm of choice within real material simulations, we then show how to reduce a sign problem, allowing us to address the unusual criticality found in layered organic superconductors. With high-T c superconductivity as example, we further discuss how to ensure ergodicity of the CT-HYB solver in the context of broken symmetries. Finally, algorithmic optimizations of CT-HYB are presented and combined, leading to speedups of up to 500 within the context of real material simulations. // Résumé: Initialement conçue pour traiter les fortes corrélations électroniques dans des hamiltoniens modèles, la théorie du champ moyen dynamique (DMFT) est devenue, en combinaison avec la théorie de la densité fonctionnelle (DFT), un outil puissant pour la simulation de matériaux réels. Au cœur de la DMFT se trouve la solution d'un modèle d'impureté quantique. Seulement les solutionneurs d'impureté Monte Carlo en temps continu (CT-QMC) donnent des solutions exactes. En même temps, ces solutionneurs sont plutôt complexes et gourmands en temps de calcul. Le but de cette thèse est donc d'améliorer les solutionneurs d'impureté CT-QMC. Après une courte introduction à la DMFT et à ses extensions pour les amas, on commence par une revue de deux des solutionneurs CT-QMC, celui en développement d'interaction ou de "Rubtsov" (CT-INT) et celui en développement d'hybridation (CT-HYB). Mettant l'accent sur le dernier, qui est l'algorithme de choix dans le cadre des matériaux réels, on montre alors comment réduire un problème de signe, nous permettant ainsi de traiter la criticalité inhabituelle des organiques en couche. Avec la supraconductivité à haute température critique comme exemple, on discute ensuite comment assurer l'ergodicité du solutionneur CT-HYB dans le cadre des symétries brisées. Finalement, des optimisations algorithmiques sont présentées et combinées, amenant à des accélérations allant jusqu'à un facteur de 500 dans le contexte des matériaux réels.

•Jan 1, 2014•1 citation

Continuous-time auxiliary field Monte Carlo for quantum impurity models

Emanuel Gull, Olivier Parcollet, Matthias Troyer

We present a continuous-time Monte Carlo method for quantum impurity models, which combines a weak-coupling expansion with an auxiliary-field decomposition. The method is considerably more efficient than Hirsch-Fye and free of time discretization errors, and is particularly useful as impurity solver in large cluster dynamical mean-field theory (DMFT) calculations.

•Nov 20, 2012•122 citations

Continuous-time auxiliary-field Monte Carlo for quantum impurity models

Emanuel Gull, Philipp Werner, Olivier Parcollet +1 more

We present a continuous-time Monte Carlo method for quantum impurity models, which combines a weak-coupling expansion with an auxiliary-field decomposition. The method is considerably more efficient than Hirsch-Fye and free of time discretization errors, and is particularly useful as impurity solver in large cluster dynamical mean-field theory (DMFT) calculations.

arXiv:0802.3222•Europhysics Letters (EPL)•May 27, 2008•265 citations
An acceleration of continuous time quantum Monte Carlo (CTQMC) methods is a potentially interesting branch of work as they are matchless as impurity solvers of a dynamical mean field theory (DMFT) approach for the description of strongly correlated systems. The inversion of the k\ifmmode×\else\texttimes\fikk\ifmmode\times\else\texttimes\fi{}kk\ifmmode×\else\texttimes\fik matrix with k2{k}^{2}k2 operations given by the diagram expansion order kkk in the CTQMC fast update and the multiplication of the k\ifmmode×\else\texttimes\fikk\ifmmode\times\else\texttimes\fi{}kk\ifmmode×\else\texttimes\fik matrix, and the noninteracting properties with k\ifmmode×\else\texttimes\fi\ensuremathωnmaxm\ensuremath−1k\ifmmode\times\else\texttimes\fi{}{\ensuremath{\omega}}_{{n}_{\text{max}}}^{m\ensuremath{-}1}k\ifmmode×\else\texttimes\fi\ensuremathωnmax​m\ensuremath−1​ operations to measure the mmm-point correlators, are computationally time consuming. Here we propose the CTQMC method in combination with a machine learning technique, which eliminates the k\ifmmode×\else\texttimes\fi\ensuremathωnmaxk\ifmmode\times\else\texttimes\fi{}{\ensuremath{\omega}}_{{n}_{\text{max}}}k\ifmmode×\else\texttimes\fi\ensuremathωnmax​​ and k\ifmmode×\else\texttimes\fi\ensuremathωnmax3k\ifmmode\times\else\texttimes\fi{}{\ensuremath{\omega}}_{{n}_{\text{max}}}^{3}k\ifmmode×\else\texttimes\fi\ensuremathωnmax​3​ operations for the two-point impurity Green's functions G\ensuremathσ(i\ensuremathωn){G}_{\ensuremath{\sigma}}(i{\ensuremath{\omega}}_{n})G\ensuremathσ​(i\ensuremathωn​) and four-point vertices \ensuremathχ\ensuremathσ,\ensuremathσ‾(i\ensuremathωn1,i\ensuremathωn2,i\ensuremathωn3,i\ensuremathωn4){\ensuremath{\chi}}_{\ensuremath{\sigma},\overline{\ensuremath{\sigma}}}(i{\ensuremath{\omega}}_{n{}_{1}},i{\ensuremath{\omega}}_{n{}_{2}},i{\ensuremath{\omega}}_{n{}_{3}},i{\ensuremath{\omega}}_{n{}_{4}})\ensuremathχ\ensuremathσ,\ensuremathσ​​(i\ensuremathωn1​​,i\ensuremathωn2​​,i\ensuremathωn3​​,i\ensuremathωn4​​), respectively. This method not only predicts the accurate physical properties at low temperature, but also dramatically decreases the computational times of \ensuremathχ\ensuremathσ,\ensuremathσ‾(i\ensuremathωn1,i\ensuremathωn2,i\ensuremathωn3,i\ensuremathωn4){\ensuremath{\chi}}_{\ensuremath{\sigma},\overline{\ensuremath{\sigma}}}(i{\ensuremath{\omega}}_{n{}_{1}},i{\ensuremath{\omega}}_{n{}_{2}},i{\ensuremath{\omega}}_{n{}_{3}},i{\ensuremath{\omega}}_{n{}_{4}})\ensuremathχ\ensuremathσ,\ensuremathσ​​(i\ensuremathωn1​​,i\ensuremathωn2​​,i\ensuremathωn3​​,i\ensuremathωn4​​) for the nonlocal extension of DMFT approximation.
The acceleration of exact continuous time quantum Monte Carlo (CTQMC) approaches in multi-site or multi-orbital systems is extremely interesting work, because these approaches are very time-consuming in terms of numerical computation and might account for the nature of exotic behaviors such as high-temperature superconductivity and Mott insulator behavior observed in the strongly correlated materials. We extend the recently developed interaction-expansion CTQMC method in combination with a machine learning (CTQMC+ML) approach for the single-site and single-orbital systems to multi-site and multi-orbital ones. This method can be applied to explore the nonlocal correlation effects in lattice models and to study the electronic structure of real materials via an ab-initio density functional theory plus dynamical mean field theory approach. We find that our CTQMC+ML method for multi-site (and multi-orbital) systems accurately predicts the impurity Green’s function with less computational time than the CTQMC approaches, as in the case of the single-site and single-orbital version.
Dynamical mean-field theory (DMFT) is one of the most widely-used methods to treat accurately electron correlation effects in ab-initio real material calculations. Many modern large-scale implementations of DMFT in electronic structure codes involve solving a quantum impurity model with a continuous-time quantum Monte Carlo (CT-QMC) solver [Rubtsov et al., Phys. Rev. B 72, 035122 (2005); Werner et al., Phys. Rev. Lett. 97, 076405 (2006); Werner and Millis, Phys. Rev. B 74, 155107 (2006); Gull et al., Rev. Mod. Phys. 83, 349 (2011)]. The main advantage of CT-QMC is that, unlike standard quantum Monte Carlo approaches, it is able to generate the local Green's functions $G(\ensuremath{\tau})$ of the correlated system on an arbitrarily fine imaginary time $\ensuremath{\tau}$ grid, and is free of any systematic errors. In this work, we extend a hybrid QMC solver proposed by Khatami et al. [Phys. Rev. E 81, 056703 (2010)] and Rost et al. [Phys. Rev. E 87, 053305 (2013)] to a multiorbital context. This has the advantage of enabling impurity solver QMC calculations to scale linearly with inverse temperature $\ensuremath{\beta}$, and permit its application to ddd- and fff-band materials. In addition, we present a Green's-function processing scheme which generates accurate quasicontinuous imaginary time solutions of the impurity problem which overcome errors inherent to standard QMC approaches. This solver and processing scheme are incorporated into a full DFT+DMFT calculation using the castep DFT code [Clark et al., Z. Kristallogr. 220, 567 (2005)]. Benchmark calculations for SrVO3{\mathrm{SrVO}}_{3}SrVO3​ properties are presented. The computational efficiency of this method is also demonstrated.
Continuous-time Quantum Monte Carlo (CTQMC) impurity solvers are a numerically exact, but expensive, option for solving the Dynamical Mean Field Theory (DMFT) equations, with statistical noise from undersampling presenting a major challenge to low temperature studies. This noise is especially egregious within the sampled self-energy, which can destabilize the self-consistency loop upon insertion into the Dyson equation. In this work, we introduce a neural denoiser for fermionic imaginary time Green's functions based on the diffusion model architecture trained on synthetic data. The network predicts positive and normalized auxiliary spectral functions consistent with the noisy imaginary axis CTQMC data, which is then refined via autodifferentiation to match the measured self-energy, its first two high-frequency moments, and impurity occupation similar to the Constrained Residual Minimization (CRM) method. This leads to Green's functions and self-energies that are causal by design. Through this denoising framework we are able to reduce the required CTQMC measurement steps by a factor of 10-100. We benchmark our method with a half-filled Hubbard model on a Bethe lattice, a two-patch Dynamical Cluster Approximation (DCA) of the two-dimensional Hubbard model, and the SrVO3_33​ system. We reproduce the Mott transition upon increasing Coulomb interaction for the half-filled Hubbard model with only 5000 CTQMC measurements, the pseudogap transition for the DCA two-dimensional Hubbard model with only 5$\times101010^5$ measurements, and match the spectral function of SrVO3_33​ with 50 times fewer measurements than the reference. Since we generate auxiliary spectral functions, we recover the approximate spectral functions of the systems for free without performing analytic continuation.
Abstract: Originally designed for the study of strong electronic correlations in model Hamiltonians, dynamical mean field theory (DMFT) has become, in combination with density functional theory (DFT), a powerful tool for ab initio simulations of real materials. At the heart of DMFT lies the solution of a quantum impurity problem. While only the continuous-time quantum Monte Carlo (CT-QMC) impurity solvers yield (statistically) exact solutions of a general impurity problem, they are quite complex and computationally expensive. Hence, in this thesis we are interested in improving the CT-QMC impurity solvers. After a short introduction to DMFT and its cluster extensions, we begin by reviewing two of the CT-QMC impurity solvers, the interaction expansion or “Rubtsov” solver (CT- INT) and the hybridization expansion solver (CT-HYB). Focussing on the latter, which is the algorithm of choice within real material simulations, we then show how to reduce a sign problem, allowing us to address the unusual criticality found in layered organic superconductors. With high-T c superconductivity as example, we further discuss how to ensure ergodicity of the CT-HYB solver in the context of broken symmetries. Finally, algorithmic optimizations of CT-HYB are presented and combined, leading to speedups of up to 500 within the context of real material simulations. // Résumé: Initialement conçue pour traiter les fortes corrélations électroniques dans des hamiltoniens modèles, la théorie du champ moyen dynamique (DMFT) est devenue, en combinaison avec la théorie de la densité fonctionnelle (DFT), un outil puissant pour la simulation de matériaux réels. Au cœur de la DMFT se trouve la solution d'un modèle d'impureté quantique. Seulement les solutionneurs d'impureté Monte Carlo en temps continu (CT-QMC) donnent des solutions exactes. En même temps, ces solutionneurs sont plutôt complexes et gourmands en temps de calcul. Le but de cette thèse est donc d'améliorer les solutionneurs d'impureté CT-QMC. Après une courte introduction à la DMFT et à ses extensions pour les amas, on commence par une revue de deux des solutionneurs CT-QMC, celui en développement d'interaction ou de "Rubtsov" (CT-INT) et celui en développement d'hybridation (CT-HYB). Mettant l'accent sur le dernier, qui est l'algorithme de choix dans le cadre des matériaux réels, on montre alors comment réduire un problème de signe, nous permettant ainsi de traiter la criticalité inhabituelle des organiques en couche. Avec la supraconductivité à haute température critique comme exemple, on discute ensuite comment assurer l'ergodicité du solutionneur CT-HYB dans le cadre des symétries brisées. Finalement, des optimisations algorithmiques sont présentées et combinées, amenant à des accélérations allant jusqu'à un facteur de 500 dans le contexte des matériaux réels.