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:2204.06769 · Communications in Theoretical Physics

Learning topological defects formation with neural networks in a quantum phase transition

Han-Qing Shi, Hai-Qing Zhang

Published Mar 11, 2024• 1 citation
arXivHTMLPDF

Abstract

Abstract Neural networks possess formidable representational power, rendering them invaluable in solving complex quantum many-body systems. While they excel at analyzing static solutions, nonequilibrium processes, including critical dynamics during a quantum phase transition, pose a greater challenge for neural networks. To address this, we utilize neural networks and machine learning algorithms to investigate time evolutions, universal statistics, and correlations of topological defects in a one-dimensional transverse-field quantum Ising model. Specifically, our analysis involves computing the energy of the system during a quantum phase transition following a linear quench of the transverse magnetic field strength. The excitation energies satisfy a power-law relation to the quench rate, indicating a proportional relationship between the excitation energy and the kink numbers. Moreover, we establish a universal power-law relationship between the first three cumulants of the kink numbers and the quench rate, indicating a binomial distribution of the kinks. Finally, the normalized kink-kink correlations are also investigated and it is found that the numerical values are consistent with the analytic formula.

DOI
10.1088/1572-9494/ad3227

Related papers

Universal Statistics of Topological Defects Formed in a Quantum Phase Transition

Adolfo del Campo

When a quantum phase transition is crossed in finite time, critical slowing down leads to the breakdown of adiabatic dynamics and the formation of topological defects. The average density of defects scales with the quench rate following a universal power law predicted by the Kibble-Zurek mechanism. We analyze the full counting statistics of kinks and report the exact kink number distribution in the transverse-field quantum Ising model. Kink statistics is described by the Poisson binomial distribution with all cumulants exhibiting a universal power law scaling with the quench rate. In the absence of finite-size effects, the distribution approaches a normal one, a feature that is expected to apply broadly in systems described by the Kibble-Zurek mechanism.

arXiv:1806.10646•Physical Review Letters•Nov 14, 2018•97 citations

Statistics of topological defects across a phase transition in a digital superconducting quantum processor

Oriel Kiss, Daniil Teplitskiy, Michele Grossi +1 more

Abstract When a quantum phase transition is crossed within a finite time, critical slowing down disrupts adiabatic dynamics, resulting in the formation of topological defects. The average density of these defects scales with the quench rate, adhering to a universal power law as predicted by the Kibble–Zurek mechanism (KZM). In this study, we aim to investigate the counting statistics of kink density in the 1D transverse-field quantum Ising model. We demonstrate on multiple quantum processing units up to 100 qubits, that higher-order cumulants follow a universal power law scaling as a function of the quench time. We also show the breakdown of the KZM for short quenches for finite-size systems. Tensor network simulations corroborate our quantum simulation results for bigger systems not in the asymptotic limit.

arXiv:2410.06250•Quantum Science and Technology•Jun 2, 2025•5 citations

Learning Rényi entropy in neural network quantum states

Han-Qing Shi, Hai-Qing Zhang

Abstract We study the Rényi entropy in a one-dimensional transverse-field quantum Ising model by virtue of the machine learning methods with neural networks. The Rényi entropy is investigated both in the static ground state and in the dynamical state after a linear quench, by employing a swapping operator acting on the states prepared from the neural networks. In the ground state, the behavior of Rényi entropy can unveil the critical point from paramagnetic to ferromagnetic phase. At the critical point, the relation between the Rényi entropy and the subsystem size from the machine learning method satisfies the predictions from conformal field theory. In the quenching dynamics, coherent oscillations of the Rényi entropy are observed after the end of the quench. These oscillations have universal frequencies stem from the superpositions of the excited states. More importantly, an asymptotic scaling of the reduced Rényi entropy is observed, implying a novel length scale away from the critical point.

Physica Scripta•Oct 21, 2026•0 citations

Probing criticality in quantum spin chains with neural networks

A. V. Berezutskii, M Beketov, D Yudin +2 more

Abstract The numerical emulation of quantum systems often requires an exponential number of degrees of freedom which translates to a computational bottleneck. Methods of machine learning have been used in adjacent fields for effective feature extraction and dimensionality reduction of high-dimensional datasets. Recent studies have revealed that neural networks are further suitable for the determination of macroscopic phases of matter and associated phase transitions as well as efficient quantum state representation. In this work, we address quantum phase transitions in quantum spin chains, namely the transverse field Ising chain and the anisotropic XY chain, and show that even neural networks with no hidden layers can be effectively trained to distinguish between magnetically ordered and disordered phases. Our neural network acts to predict the corresponding crossovers finite-size systems undergo. Our results extend to a wide class of interacting quantum many-body systems and illustrate the wide applicability of neural networks to many-body quantum physics.

arXiv:2005.02104•Journal of Physics Complexity•Jul 28, 2020•8 citations

Probing criticality in quantum spin chains with neural networks

Berezutskii, Aleksander, Beketov, Maksim, Yudin, Dmitry +2 more

The numerical emulation of quantum systems often requires an exponential number of degrees of freedom which translates to a computational bottleneck. Methods of machine learning have been used in adjacent fields for effective feature extraction and dimensionality reduction of high-dimensional datasets. Recent studies have revealed that neural networks are further suitable for the determination of macroscopic phases of matter and associated phase transitions as well as efficient quantum state representation. In this work, we address quantum phase transitions in quantum spin chains, namely the transverse field Ising chain and the anisotropic XY chain, and show that even neural networks with no hidden layers can be effectively trained to distinguish between magnetically ordered and disordered phases. Our neural network acts to predict the corresponding crossovers finite-size systems undergo. Our results extend to a wide class of interacting quantum many-body systems and illustrate the wide applicability of neural networks to many-body quantum physics.

•Jan 1, 2020•7 citations

Driving a quantum phase transition at an arbitrary rate: Exact solution of the transverse-field Ising model

András Grabarits, Federico Balducci, Adolfo del Campo

We study the crossing of the quantum phase transition in the transverse-field Ising model after modulating the magnetic field at an arbitrary rate, exploring the critical dynamics from the slow to the sudden-quench regime. We do so by analyzing the defect density, the complete kink-number distribution, and its cumulants upon completion of a linearized quench. Our analysis relies on the diagonalization of the model using the standard Jordan-Wigner and Fourier transformations, along with the exact solution of the time evolution in each mode in terms of parabolic cylinder functions. The free-fermion nature of the problem dictates that the kink-number distribution is associated with independent and distinguishable Bernoulli variables, each with a success probability ${p}_{k}$. We employ a combination of convergent and asymptotic series expansions to characterize ${p}_{k}$ without restrictions on the driving rate. When the latter is approximated by the Landau-Zener formula, the kink density is described by the Kibble-Zurek mechanism, and higher-order cumulants follow a universal power-law behavior, as recently predicted theoretically and verified in the laboratory. By contrast, for moderate and sudden-driving protocols, the cumulants exhibit a nonuniversal behavior that is not monotonic on the driving rate and changes qualitatively with the cumulant order. The kink-number statistics remain sub-Poissonian for any driving rate, as revealed by an analysis of the cumulant ratios that vary nonmonotonically from the sudden- to the slow-driving regime. Due to the determination of ${p}_{k}$ for an arbitrary rate, our study provides a complete analytical understanding of kink-number statistics from the slow-driving regime to the sudden-quench limit.

arXiv:2501.10478•Physical Review A•Apr 16, 2025•7 citations
Abstract Neural networks possess formidable representational power, rendering them invaluable in solving complex quantum many-body systems. While they excel at analyzing static solutions, nonequilibrium processes, including critical dynamics during a quantum phase transition, pose a greater challenge for neural networks. To address this, we utilize neural networks and machine learning algorithms to investigate time evolutions, universal statistics, and correlations of topological defects in a one-dimensional transverse-field quantum Ising model. Specifically, our analysis involves computing the energy of the system during a quantum phase transition following a linear quench of the transverse magnetic field strength. The excitation energies satisfy a power-law relation to the quench rate, indicating a proportional relationship between the excitation energy and the kink numbers. Moreover, we establish a universal power-law relationship between the first three cumulants of the kink numbers and the quench rate, indicating a binomial distribution of the kinks. Finally, the normalized kink-kink correlations are also investigated and it is found that the numerical values are consistent with the analytic formula.
When a quantum phase transition is crossed in finite time, critical slowing down leads to the breakdown of adiabatic dynamics and the formation of topological defects. The average density of defects scales with the quench rate following a universal power law predicted by the Kibble-Zurek mechanism. We analyze the full counting statistics of kinks and report the exact kink number distribution in the transverse-field quantum Ising model. Kink statistics is described by the Poisson binomial distribution with all cumulants exhibiting a universal power law scaling with the quench rate. In the absence of finite-size effects, the distribution approaches a normal one, a feature that is expected to apply broadly in systems described by the Kibble-Zurek mechanism.
Abstract When a quantum phase transition is crossed within a finite time, critical slowing down disrupts adiabatic dynamics, resulting in the formation of topological defects. The average density of these defects scales with the quench rate, adhering to a universal power law as predicted by the Kibble–Zurek mechanism (KZM). In this study, we aim to investigate the counting statistics of kink density in the 1D transverse-field quantum Ising model. We demonstrate on multiple quantum processing units up to 100 qubits, that higher-order cumulants follow a universal power law scaling as a function of the quench time. We also show the breakdown of the KZM for short quenches for finite-size systems. Tensor network simulations corroborate our quantum simulation results for bigger systems not in the asymptotic limit.
Abstract We study the Rényi entropy in a one-dimensional transverse-field quantum Ising model by virtue of the machine learning methods with neural networks. The Rényi entropy is investigated both in the static ground state and in the dynamical state after a linear quench, by employing a swapping operator acting on the states prepared from the neural networks. In the ground state, the behavior of Rényi entropy can unveil the critical point from paramagnetic to ferromagnetic phase. At the critical point, the relation between the Rényi entropy and the subsystem size from the machine learning method satisfies the predictions from conformal field theory. In the quenching dynamics, coherent oscillations of the Rényi entropy are observed after the end of the quench. These oscillations have universal frequencies stem from the superpositions of the excited states. More importantly, an asymptotic scaling of the reduced Rényi entropy is observed, implying a novel length scale away from the critical point.
Abstract The numerical emulation of quantum systems often requires an exponential number of degrees of freedom which translates to a computational bottleneck. Methods of machine learning have been used in adjacent fields for effective feature extraction and dimensionality reduction of high-dimensional datasets. Recent studies have revealed that neural networks are further suitable for the determination of macroscopic phases of matter and associated phase transitions as well as efficient quantum state representation. In this work, we address quantum phase transitions in quantum spin chains, namely the transverse field Ising chain and the anisotropic XY chain, and show that even neural networks with no hidden layers can be effectively trained to distinguish between magnetically ordered and disordered phases. Our neural network acts to predict the corresponding crossovers finite-size systems undergo. Our results extend to a wide class of interacting quantum many-body systems and illustrate the wide applicability of neural networks to many-body quantum physics.
The numerical emulation of quantum systems often requires an exponential number of degrees of freedom which translates to a computational bottleneck. Methods of machine learning have been used in adjacent fields for effective feature extraction and dimensionality reduction of high-dimensional datasets. Recent studies have revealed that neural networks are further suitable for the determination of macroscopic phases of matter and associated phase transitions as well as efficient quantum state representation. In this work, we address quantum phase transitions in quantum spin chains, namely the transverse field Ising chain and the anisotropic XY chain, and show that even neural networks with no hidden layers can be effectively trained to distinguish between magnetically ordered and disordered phases. Our neural network acts to predict the corresponding crossovers finite-size systems undergo. Our results extend to a wide class of interacting quantum many-body systems and illustrate the wide applicability of neural networks to many-body quantum physics.
We study the crossing of the quantum phase transition in the transverse-field Ising model after modulating the magnetic field at an arbitrary rate, exploring the critical dynamics from the slow to the sudden-quench regime. We do so by analyzing the defect density, the complete kink-number distribution, and its cumulants upon completion of a linearized quench. Our analysis relies on the diagonalization of the model using the standard Jordan-Wigner and Fourier transformations, along with the exact solution of the time evolution in each mode in terms of parabolic cylinder functions. The free-fermion nature of the problem dictates that the kink-number distribution is associated with independent and distinguishable Bernoulli variables, each with a success probability pk{p}_{k}pk​. We employ a combination of convergent and asymptotic series expansions to characterize pk{p}_{k}pk​ without restrictions on the driving rate. When the latter is approximated by the Landau-Zener formula, the kink density is described by the Kibble-Zurek mechanism, and higher-order cumulants follow a universal power-law behavior, as recently predicted theoretically and verified in the laboratory. By contrast, for moderate and sudden-driving protocols, the cumulants exhibit a nonuniversal behavior that is not monotonic on the driving rate and changes qualitatively with the cumulant order. The kink-number statistics remain sub-Poissonian for any driving rate, as revealed by an analysis of the cumulant ratios that vary nonmonotonically from the sudden- to the slow-driving regime. Due to the determination of pk{p}_{k}pk​ for an arbitrary rate, our study provides a complete analytical understanding of kink-number statistics from the slow-driving regime to the sudden-quench limit.