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:0807.0682 · Physical Review B

Loop updates for variational and projector quantum Monte Carlo simulations in the valence-bond basis

Anders W. Sandvik, Hans Gerd Evertz

Published Jul 8, 2010• 144 citations
arXivHTMLPDF

Abstract

We show how efficient loop updates, originally developed for Monte Carlo simulations of quantum spin systems at finite temperature, can be combined with a ground-state projector scheme and variational calculations in the valence-bond basis. The methods are formulated in a combined space of spin $z$ components and valence bonds. Compared to schemes formulated purely in the valence-bond basis, the computational effort is reduced from up to $O({N}^{2})$ to $O(N)$ for variational calculations, where $N$ is the system size, and from $O({m}^{2})$ to $O(m)$ for projector simulations, where $m⪢N$ is the projection power. These improvements enable access to ground states of significantly larger lattices than previously. We demonstrate the efficiency of the approach by calculating the sublattice magnetization ${M}_{s}$ of the two-dimensional Heisenberg model to high precision, using systems with up to $256\ifmmode\times\else\texttimes\fi{}256$ spins. Extrapolating the results to the thermodynamic limit gives ${M}_{s}=0.30743(1)$. We also discuss optimized variational amplitude-product states, which were used as trial states in the projector simulations, and compare results of projecting different types of trial states.

DOI
10.1103/physrevb.82.024407

Related papers

Monte Carlo Simulations of Quantum Spin Systems in the Valence Bond Basis

Anders W. Sandvik, K. S. D. Beach

We discuss a projector Monte Carlo method for quantum spin models formulated in the valence bond basis, using the S=1/2 Heisenberg antiferromagnet as an example. Its singlet ground state can be projected out of an arbitrary basis state as the trial state, but a more rapid convergence can be obtained using a good variational state. As an alternative to first carrying out a time consuming variational Monte Carlo calculation, we show that a very good trial state can be generated in an iterative fashion in the course of the simulation itself. We also show how the properties of the valence bond basis enable calculations of quantities that are difficult to obtain with the standard basis of Sz eigenstates. In particular, we discuss quantities involving finite-momentum states in the triplet sector, such as the dispersion relation and the spectral weight of the lowest triplet.

arXiv:0704.1469•Apr 11, 2007•6 citations

Variational ground states of two-dimensional antiferromagnets in the valence bond basis

Jie Lou, Anders W. Sandvik

We study a variational wave function for the ground state of the two-dimensional $S=1∕2$ Heisenberg antiferromagnet in the valence bond basis. The expansion coefficients are products of amplitudes $h(x,y)$ for valence bonds connecting spins separated by $(x,y)$ lattice spacings. In contrast to previous studies, in which a functional form for $h(x,y)$ was assumed, we here optimize all the amplitudes for lattices with up to $32\ifmmode\times\else\texttimes\fi{}32$ spins. We use two different schemes for optimizing the amplitudes; a Newton conjugate-gradient method and a stochastic method which requires only the signs of the first derivatives of the energy. The latter method performs significantly better. The energy for large systems deviates by only $\ensuremath{\approx}0.06%$ from its exact value (calculated using unbiased quantum Monte Carlo simulations). The spin correlations are also well reproduced, falling $\ensuremath{\approx}2%$ below the exact ones at long distances (corresponding to an $\ensuremath{\approx}1%$ underestimation of the sublattice magnetization). The amplitudes $h(r)$ for valence bonds of long length $r$ decay as ${r}^{\ensuremath{-}3}$. We also discuss some results for small frustrated lattices.

arXiv:cond-mat/0605034•Physical Review B•Sep 25, 2007•39 citations

Ground State Projection of Quantum Spin Systems in the Valence-Bond Basis

Anders W. Sandvik

A Monte Carlo method for quantum spin systems is formulated in the valence-bond basis. The nonorthogonality allows for an efficient importance-sampled projection of the ground state out of an arbitrary state. The method provides access to resonating valence-bond physics, enables a direct estimator for the singlet-triplet gap, and extends the class of models that can be studied without negative-sign problems. As a demonstration, the valence-bond distribution in the ground state of the 2D Heisenberg antiferromagnet is calculated. Generalizations of the method to fermion systems are also discussed.

arXiv:cond-mat/0509558•Physical Review Letters•Nov 8, 2005•106 citations

Variational Monte Carlo simulations using tensor-product projected states

Olga Sikora, Hsueh-Wen Chang, Chung-Pin Chou +2 more

We propose an efficient numerical method, which combines the advantages of recently developed tensor-network based methods and standard trial wave functions, to study the ground-state properties of quantum many-body systems. In this approach, we apply a projector in the form of a tensor-product operator to an input wave function, such as a Jastrow-type or Hartree-Fock wave function, and optimize the tensor elements via variational Monte Carlo. The entanglement already contained in the input wave function can considerably reduce the bond dimensions compared to the regular tensor-product state representation. In particular, this allows us to also represent states that do not obey the area law of entanglement entropy. In addition, for fermionic systems, the fermion sign structure can be encoded in the input wave function. We show that the optimized states provide good approximations of the ground-state energy and correlation functions in the cases of two-dimensional bosonic and fermonic systems.

arXiv:1407.4107•Physical Review B•Apr 8, 2015•18 citations

Monte Carlo approach to phase transitions in quantum systems

Naoki Kawashima, Yasuyuki Kato

We propose algorithms of the path-ingetral-based quantum Monte Carlo simulation, which is otherwise prohibitively slow. While the basic idea is the loop-cluster update, there are some important 'tricks' that are vital to make the simulation practical. In the present paper, we show two such techniques and their successful applications to the two-dimensional SU( N ) Heisenberg model and the three-dimensional Bose Hubbard model. In the former, we obtain a new type of the valence-bond-solid state for the two-boson representation, while in the latter we equillibrate a system of which the size is comparable to a typical experiment of optical lattices.

Journal of Physics Conference Series•Jan 7, 2009•2 citations

Valence bond distribution and correlation in bipartite Heisenberg antiferromagnets

David R. Schwandt, Fabien Alet, Masaki Oshikawa

Every singlet state of a quantum spin-$1/2$ system can be decomposed into a linear combination of valence bond basis states. The range of valence bonds within this linear combination as well as the correlations between them can reveal the nature of the singlet state and are key ingredients in variational calculations. In this work, we study the bipartite valence bond distributions and their correlations within the ground state of the Heisenberg antiferromagnet on bipartite lattices. In terms of field theory, this problem can be mapped to correlation functions near a boundary. In dimension $d\ensuremath{\ge}2$, a nonlinear $\ensuremath{\sigma}$ model analysis reveals that at long distances the probability distribution $P(\mathbf{r})$ of valence bond lengths decays as ${|r|}^{\ensuremath{-}d\ensuremath{-}1}$ and that valence bonds are uncorrelated. By a bosonization analysis, we also obtain $P(\mathbf{r})\ensuremath{\propto}{|r|}^{\ensuremath{-}d\ensuremath{-}1}$ in $d=1$ despite the different mechanism. On the other hand, we find that correlations between valence bonds are important even at large distances in $d=1$, in stark contrast to $d\ensuremath{\ge}2$. The analytical results are confirmed by high-precision quantum Monte Carlo simulations in $d=1$, 2, and 3. We develop a single-projection loop variant of the valence bond projection algorithm, which is well designed to compute valence bond probabilities and for which we provide algorithmic details.

arXiv:1311.3975•Physical Review B•Mar 19, 2014•4 citations
We show how efficient loop updates, originally developed for Monte Carlo simulations of quantum spin systems at finite temperature, can be combined with a ground-state projector scheme and variational calculations in the valence-bond basis. The methods are formulated in a combined space of spin zzz components and valence bonds. Compared to schemes formulated purely in the valence-bond basis, the computational effort is reduced from up to O(N2)O({N}^{2})O(N2) to O(N)O(N)O(N) for variational calculations, where NNN is the system size, and from O(m2)O({m}^{2})O(m2) to O(m)O(m)O(m) for projector simulations, where m⪢Nm⪢Nm⪢N is the projection power. These improvements enable access to ground states of significantly larger lattices than previously. We demonstrate the efficiency of the approach by calculating the sublattice magnetization Ms{M}_{s}Ms​ of the two-dimensional Heisenberg model to high precision, using systems with up to 256\ifmmode×\else\texttimes\fi256256\ifmmode\times\else\texttimes\fi{}256256\ifmmode×\else\texttimes\fi256 spins. Extrapolating the results to the thermodynamic limit gives Ms=0.30743(1){M}_{s}=0.30743(1)Ms​=0.30743(1). We also discuss optimized variational amplitude-product states, which were used as trial states in the projector simulations, and compare results of projecting different types of trial states.
We discuss a projector Monte Carlo method for quantum spin models formulated in the valence bond basis, using the S=1/2 Heisenberg antiferromagnet as an example. Its singlet ground state can be projected out of an arbitrary basis state as the trial state, but a more rapid convergence can be obtained using a good variational state. As an alternative to first carrying out a time consuming variational Monte Carlo calculation, we show that a very good trial state can be generated in an iterative fashion in the course of the simulation itself. We also show how the properties of the valence bond basis enable calculations of quantities that are difficult to obtain with the standard basis of Sz eigenstates. In particular, we discuss quantities involving finite-momentum states in the triplet sector, such as the dispersion relation and the spectral weight of the lowest triplet.
We study a variational wave function for the ground state of the two-dimensional S=1∕2S=1∕2S=1∕2 Heisenberg antiferromagnet in the valence bond basis. The expansion coefficients are products of amplitudes h(x,y)h(x,y)h(x,y) for valence bonds connecting spins separated by (x,y)(x,y)(x,y) lattice spacings. In contrast to previous studies, in which a functional form for h(x,y)h(x,y)h(x,y) was assumed, we here optimize all the amplitudes for lattices with up to $32\ifmmode\times\else\texttimes\fi{}32$ spins. We use two different schemes for optimizing the amplitudes; a Newton conjugate-gradient method and a stochastic method which requires only the signs of the first derivatives of the energy. The latter method performs significantly better. The energy for large systems deviates by only $\ensuremath{\approx}0.06%$ from its exact value (calculated using unbiased quantum Monte Carlo simulations). The spin correlations are also well reproduced, falling $\ensuremath{\approx}2%$ below the exact ones at long distances (corresponding to an $\ensuremath{\approx}1%$ underestimation of the sublattice magnetization). The amplitudes h(r)h(r)h(r) for valence bonds of long length rrr decay as ${r}^{\ensuremath{-}3}$. We also discuss some results for small frustrated lattices.
A Monte Carlo method for quantum spin systems is formulated in the valence-bond basis. The nonorthogonality allows for an efficient importance-sampled projection of the ground state out of an arbitrary state. The method provides access to resonating valence-bond physics, enables a direct estimator for the singlet-triplet gap, and extends the class of models that can be studied without negative-sign problems. As a demonstration, the valence-bond distribution in the ground state of the 2D Heisenberg antiferromagnet is calculated. Generalizations of the method to fermion systems are also discussed.
We propose an efficient numerical method, which combines the advantages of recently developed tensor-network based methods and standard trial wave functions, to study the ground-state properties of quantum many-body systems. In this approach, we apply a projector in the form of a tensor-product operator to an input wave function, such as a Jastrow-type or Hartree-Fock wave function, and optimize the tensor elements via variational Monte Carlo. The entanglement already contained in the input wave function can considerably reduce the bond dimensions compared to the regular tensor-product state representation. In particular, this allows us to also represent states that do not obey the area law of entanglement entropy. In addition, for fermionic systems, the fermion sign structure can be encoded in the input wave function. We show that the optimized states provide good approximations of the ground-state energy and correlation functions in the cases of two-dimensional bosonic and fermonic systems.
We propose algorithms of the path-ingetral-based quantum Monte Carlo simulation, which is otherwise prohibitively slow. While the basic idea is the loop-cluster update, there are some important 'tricks' that are vital to make the simulation practical. In the present paper, we show two such techniques and their successful applications to the two-dimensional SU( N ) Heisenberg model and the three-dimensional Bose Hubbard model. In the former, we obtain a new type of the valence-bond-solid state for the two-boson representation, while in the latter we equillibrate a system of which the size is comparable to a typical experiment of optical lattices.
Every singlet state of a quantum spin-1/21/21/2 system can be decomposed into a linear combination of valence bond basis states. The range of valence bonds within this linear combination as well as the correlations between them can reveal the nature of the singlet state and are key ingredients in variational calculations. In this work, we study the bipartite valence bond distributions and their correlations within the ground state of the Heisenberg antiferromagnet on bipartite lattices. In terms of field theory, this problem can be mapped to correlation functions near a boundary. In dimension $d\ensuremath{\ge}2$, a nonlinear $\ensuremath{\sigma}$ model analysis reveals that at long distances the probability distribution P(r)P(\mathbf{r})P(r) of valence bond lengths decays as ${|r|}^{\ensuremath{-}d\ensuremath{-}1}$ and that valence bonds are uncorrelated. By a bosonization analysis, we also obtain $P(\mathbf{r})\ensuremath{\propto}{|r|}^{\ensuremath{-}d\ensuremath{-}1}$ in d=1d=1d=1 despite the different mechanism. On the other hand, we find that correlations between valence bonds are important even at large distances in d=1d=1d=1, in stark contrast to $d\ensuremath{\ge}2$. The analytical results are confirmed by high-precision quantum Monte Carlo simulations in d=1d=1d=1, 2, and 3. We develop a single-projection loop variant of the valence bond projection algorithm, which is well designed to compute valence bond probabilities and for which we provide algorithmic details.