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:2402.10622 · Classical and Quantum Gravity

Towards quantum gravity with neural networks: solving the quantum Hamilton constraint of U(1) BF theory

Hanno Sahlmann, Waleed Sherif

Published Oct 8, 2024• 14 citations
arXivHTMLPDF

Abstract

Abstract In the canonical approach of loop quantum gravity, arguably the most important outstanding problem is finding and interpreting solutions to the Hamiltonian constraint. In this work, we demonstrate that methods of machine learning are in principle applicable to this problem. We consider U(1) BF theory in three dimensions, quantised with loop quantum gravity methods. In particular, we formulate a master constraint corresponding to Hamilton and Gauß constraints using loop quantum gravity methods. To make the problem amenable for numerical simulation we fix a graph and introduce a cutoff on the kinematical degrees of freedom, effectively considering U q ( 1 ) BF theory at a root of unity. We show that the neural network quantum state ansatz can be used to numerically solve the constraints efficiently and accurately. We compute expectation values and fluctuations of certain observables and compare them with exact results or exact numerical methods where possible. We also study the dependence on the cutoff.

DOI
10.1088/1361-6382/ad84af

Related papers

Towards quantum gravity with neural networks: solving quantum Hamilton constraints of 3d Euclidean gravity in the weak coupling limit

Hanno Sahlmann, Waleed Sherif

Abstract The aim of the present work is to demonstrate the applicability of machine learning techniques and in particular neural network quantum states in the context of loop quantum gravity, albeit for a simplified theory. We consider 3-dimensional Euclidean gravity in a gauge theoretical formulation in a certain weak coupling limit due to Smolin, and show that the gauge group becomes Abelian, resulting in U(1)3 BF-theory. The theory is quantised using loop quantum gravity methods. The kinematical degrees of freedom are truncated, on account of computational feasibility, by fixing a graph and deforming the algebra of the holonomies to impose a cutoff on the charge vectors. This leads to a quantum theory related to Uq(1)3 BF-theory. The effect of imposing the cutoff on the charges is examined. We also implement the quantum volume operator of 3d loop quantum gravity. Most importantly we compare two constraints for the quantum model obtained: a master constraint enforcing curvature and Gauß constraint, as well as a combination of a quantum Hamilton constraint constructed using Thiemann’s strategy and the Gauß master constraint. The two constraints are solved using the neural network quantum state ansatz, demonstrating its ability to explore models which are out of reach for exact numerical methods. The solutions spaces are quantitatively compared and although the forms of the constraints are radically different, the solutions turn out to have a surprisingly large overlap. We also investigate the behavior of the quantum volume in solutions to the constraints.

arXiv:2405.00661•Classical and Quantum Gravity•Sep 17, 2024•10 citations

Insights into physical states of loop quantum gravity using techniques from machine learning

Waleed Sherif

Loop quantum gravity offers a rich non-perturbative framework for quantum gravity, yet its dynamics remain extremely difficult to implement in practice. In the canonical formulation, physical states are defined by quantum constraints acting on spin-network states, whose graph, representation, and intertwiner data lead to a highly complex kinematical state space. In the covariant formulation, transition amplitudes are built from spin foam vertex amplitudes whose exact evaluation is computationally demanding. This dissertation develops a computational programme for the dynamics of loop quantum gravity in both its canonical and covariant formulations. The programme combines structure-preserving numerical methods with machine learning techniques, with the aim of preserving the native graph, gauge, and operator structure of the theory while making its dynamical problems accessible to controlled computation at large scale. In the canonical theory, this means using variational Monte Carlo methods and neural quantum states to approximate solutions of quantum constraints. In the covariant theory, it means using neural network surrogates to approximate spin foam vertex amplitudes. In this way, part of the dynamical problem of loop quantum gravity is brought from the purely formal domain into a regime of explicit and controlled numerical study. On the canonical side, a fixed-graph framework is developed in which graph non-enlarging Hamiltonian constraints are recast as variational problems for positive quadratic forms whose minima encode weak constraint equations. This reformulation turns the search for physical states into a ground state problem on truncated Hilbert spaces. To realise this programme in practice, neuraLQX, an open-source high-performance toolkit for variational canonical loop quantum gravity built around graph-aware Hilbert spaces, observables, and constraint operators, was developed as part of this dissertation. It uses Markov chain Monte Carlo methods, and in particular gauge-invariant sampling, stochastic reconfiguration, and quantum states parametrized by neural networks adapted to the locality of the dynamics. To demonstrate the viability of this framework, high-fidelity near-physical states are obtained for Abelian BF theory and for a weak coupling limit of LQG with structure group \(\mathrm{U}(1)^3\) in three and four dimensions. In the four-dimensional model on the complete graph \(K_5\), the analysis shows that the ordering of the vertex Hamiltonian constraint has a strong influence on which solution sectors are found variationally. The relevant operator is the vertex Hamiltonian \(\hat{H}_v\), quantized in the spirit of Thiemann's construction and with the corresponding ordering. Numerically, the weak equations associated with \(\hat{H}_v\) and \(\hat{H}_v^\dagger\) are probed through positive quadratic objectives built from \(\hat{H}_v^\dagger \hat{H}_v\) and \(\hat{H}_v \hat{H}_v^\dagger\). The analysis shows that these two searches do not merely produce different approximations to the same variational state. Instead, they select distinct solution sectors. One sector is peaked on flat connections and appears to describe a non-degenerate spatial quantum geometry. Numerical analysis shows this sector to be consistent with a discrete approximation to a Dittrich-Geiller-type vacuum. The other sector consists of states in the kernel of the volume operator. These states are non-flat and exhibit correlations across the graph, consistent with volume-zero excitations over an Ashtekar-Lewandowski-type vacuum. The relation between the two sectors is asymmetric in the sense that while the Ashtekar-Lewandowski-like states are often solutions for both orderings of the quadratic constraints, the Dittrich-Geiller-like states are not. Thus, one can conclude that the ordering of the Hamiltonian constraint affects which sectors are accessible or selected in the variational search, and it suggests that different orderings may exclude some classes of solutions. On the covariant side, this dissertation develops neural network surrogates for spin foam vertex amplitudes, with the aim of approximating the local building blocks of the path-integral dynamics. Starting with the Euclidean Barrett-Crane amplitude, the numerical problem is shown to separate naturally into two physically distinct tasks. The first is to determine which boundary spin configurations are admissible and therefore lie in the support of the amplitude. The second is to approximate the numerical value assigned to each admissible configuration by the vertex amplitude itself. The support is learned with near-perfect accuracy and generalises across cutoffs, showing that the network can infer and extrapolate the known, algorithmically verifiable \(\mathrm{SU}(2)\) recoupling conditions that determine admissibility in the Barrett-Crane case. By contrast, the amplitude values can be reproduced with high precision at fixed cutoff, but do not extrapolate reliably to larger cutoffs. This shows that the dependence of the vertex amplitude on the boundary quantum geometry is considerably more intricate than the structure of its support. The analysis also demonstrates that the representation of discrete spin data is not a merely technical modelling choice but directly affects how well the relevant quantum-geometric information is exposed to the surrogate model. This observation motivates graph-based surrogate architectures in which the spin foam boundary data and their adjacency structure are built into the model more explicitly. These architectures are then applied to the Lorentzian EPRL amplitude whereby in this case, the thesis first addresses the computational bottleneck of exact evaluation by reducing the runtime of the `sl2cfoam-next` exact evaluator sufficiently to generate large training datasets. Preliminary surrogate results show that the support, as well as the sign, of the EPRL amplitude can also be learned effectively, giving first evidence that graph-based learning can capture part of the admissibility structure of the Lorentzian theory.

•Jan 1, 2026•0 citations

Exploring the near-kernel sector of quantum reduced loop gravity

Ilkka Mäkinen, Hanno Sahlmann, Waleed Sherif

Abstract We study the near-kernel sector of the Hamiltonian constraint operator in the one-vertex model of quantum reduced loop gravity using exact diagonalisation methods and variational Monte Carlo methods with neural quantum states. The analysis is based on the symmetric Hamiltonian containing both Euclidean and Lorentzian contributions, and on the variational minimization of the positive quadratic operator $\hat{Q}=\hat{C}^2$, where $\hat{C}$ is the symmetric Hamilton constraint, in truncated Hilbert spaces with spin cutoff up to $j_{\mathrm{max}}=1001$. In the cases where exact diagonalization is possible, the eigenspace of lowest eigenvalue is one- or two-dimensional, and the eigenvalue, as well as the structure of the eigenstates depends on the cutoff. These ground states are entangled across the three edges. The variational method does not find the lowest eigenspace, but near kernel states, i.e., states with components in a number of low lying eigenspaces. For low cutoffs, they also do not factorize across the three edge degrees of freedom. At larger cutoffs, it increasingly finds near kernel states that factorize over the edges, matched with near-unit fidelity by reduced Thiemann coherent states. We discuss some indication that these states are the optimal near-kernel states if factorization is assumed.

arXiv:2605.18625•Classical and Quantum Gravity•Nov 5, 2026•2 citations

Classical Quantum Optimization with Neural Network Quantum States

Joseph Gomes, Keri A. McKiernan, Peter K. Eastman +1 more

The classical simulation of quantum systems typically requires exponential resources. Recently, the introduction of a machine learning-based wavefunction ansatz has led to the ability to solve the quantum many-body problem in regimes that had previously been intractable for existing exact numerical methods. Here, we demonstrate the utility of the variational representation of quantum states based on artificial neural networks for performing quantum optimization. We show empirically that this methodology achieves high approximation ratio solutions with polynomial classical computing resources for a range of instances of the Maximum Cut (MaxCut) problem whose solutions have been encoded into the ground state of quantum many-body systems up to and including 256 qubits.

arXiv:1910.10675•Oct 23, 2019•17 citations

Gauge Equivariant Neural Networks for Quantum Lattice Gauge Theories

Di Luo, Giuseppe Carleo, Bryan K. Clark +1 more

Gauge symmetries play a key role in physics appearing in areas such as quantum field theories of the fundamental particles and emergent degrees of freedom in quantum materials. Motivated by the desire to efficiently simulate many-body quantum systems with exact local gauge invariance, gauge equivariant neural-network quantum states are introduced, which exactly satisfy the local Hilbert space constraints necessary for the description of quantum lattice gauge theory with Z_{d} gauge group and non-Abelian Kitaev D(G) models on different geometries. Focusing on the special case of Z_{2} gauge group on a periodically identified square lattice, the equivariant architecture is analytically shown to contain the loop-gas solution as a special case. Gauge equivariant neural-network quantum states are used in combination with variational quantum Monte Carlo to obtain compact descriptions of the ground state wave function for the Z_{2} theory away from the exactly solvable limit, and to demonstrate the confining or deconfining phase transition of the Wilson loop order parameter.

arXiv:2012.05232•Physical Review Letters•Dec 30, 2021•60 citations

Neural network solutions of bosonic quantum systems in one dimension

Paulo F. Bedaque, Hersh Kumar, Andy Sheng

Neural networks have been proposed as efficient numerical wave function Ans\"atze, which can be used to variationally search a wide range of functional forms for ground-state solutions. These neural network methods are also advantageous in that more variational parameters and system degrees of freedom can be easily added. We benchmark the methodology by using neural networks to study several different integrable bosonic quantum systems in one dimension and compare our results to the exact solutions. While testing the scalability of the procedure to systems with many particles, we also introduce using symmetric function inputs to the neural network to enforce exchange symmetries of indistinguishable particles.

arXiv:2309.02352•Physical Review C•Mar 27, 2024•8 citations
Abstract In the canonical approach of loop quantum gravity, arguably the most important outstanding problem is finding and interpreting solutions to the Hamiltonian constraint. In this work, we demonstrate that methods of machine learning are in principle applicable to this problem. We consider U(1) BF theory in three dimensions, quantised with loop quantum gravity methods. In particular, we formulate a master constraint corresponding to Hamilton and Gauß constraints using loop quantum gravity methods. To make the problem amenable for numerical simulation we fix a graph and introduce a cutoff on the kinematical degrees of freedom, effectively considering U q ( 1 ) BF theory at a root of unity. We show that the neural network quantum state ansatz can be used to numerically solve the constraints efficiently and accurately. We compute expectation values and fluctuations of certain observables and compare them with exact results or exact numerical methods where possible. We also study the dependence on the cutoff.
Abstract The aim of the present work is to demonstrate the applicability of machine learning techniques and in particular neural network quantum states in the context of loop quantum gravity, albeit for a simplified theory. We consider 3-dimensional Euclidean gravity in a gauge theoretical formulation in a certain weak coupling limit due to Smolin, and show that the gauge group becomes Abelian, resulting in U(1)3 BF-theory. The theory is quantised using loop quantum gravity methods. The kinematical degrees of freedom are truncated, on account of computational feasibility, by fixing a graph and deforming the algebra of the holonomies to impose a cutoff on the charge vectors. This leads to a quantum theory related to Uq(1)3 BF-theory. The effect of imposing the cutoff on the charges is examined. We also implement the quantum volume operator of 3d loop quantum gravity. Most importantly we compare two constraints for the quantum model obtained: a master constraint enforcing curvature and Gauß constraint, as well as a combination of a quantum Hamilton constraint constructed using Thiemann’s strategy and the Gauß master constraint. The two constraints are solved using the neural network quantum state ansatz, demonstrating its ability to explore models which are out of reach for exact numerical methods. The solutions spaces are quantitatively compared and although the forms of the constraints are radically different, the solutions turn out to have a surprisingly large overlap. We also investigate the behavior of the quantum volume in solutions to the constraints.
Loop quantum gravity offers a rich non-perturbative framework for quantum gravity, yet its dynamics remain extremely difficult to implement in practice. In the canonical formulation, physical states are defined by quantum constraints acting on spin-network states, whose graph, representation, and intertwiner data lead to a highly complex kinematical state space. In the covariant formulation, transition amplitudes are built from spin foam vertex amplitudes whose exact evaluation is computationally demanding. This dissertation develops a computational programme for the dynamics of loop quantum gravity in both its canonical and covariant formulations. The programme combines structure-preserving numerical methods with machine learning techniques, with the aim of preserving the native graph, gauge, and operator structure of the theory while making its dynamical problems accessible to controlled computation at large scale. In the canonical theory, this means using variational Monte Carlo methods and neural quantum states to approximate solutions of quantum constraints. In the covariant theory, it means using neural network surrogates to approximate spin foam vertex amplitudes. In this way, part of the dynamical problem of loop quantum gravity is brought from the purely formal domain into a regime of explicit and controlled numerical study. On the canonical side, a fixed-graph framework is developed in which graph non-enlarging Hamiltonian constraints are recast as variational problems for positive quadratic forms whose minima encode weak constraint equations. This reformulation turns the search for physical states into a ground state problem on truncated Hilbert spaces. To realise this programme in practice, neuraLQX, an open-source high-performance toolkit for variational canonical loop quantum gravity built around graph-aware Hilbert spaces, observables, and constraint operators, was developed as part of this dissertation. It uses Markov chain Monte Carlo methods, and in particular gauge-invariant sampling, stochastic reconfiguration, and quantum states parametrized by neural networks adapted to the locality of the dynamics. To demonstrate the viability of this framework, high-fidelity near-physical states are obtained for Abelian BF theory and for a weak coupling limit of LQG with structure group U(1)3\mathrm{U}(1)^3U(1)3 in three and four dimensions. In the four-dimensional model on the complete graph K5K_5K5​, the analysis shows that the ordering of the vertex Hamiltonian constraint has a strong influence on which solution sectors are found variationally. The relevant operator is the vertex Hamiltonian H^v\hat{H}_vH^v​, quantized in the spirit of Thiemann's construction and with the corresponding ordering. Numerically, the weak equations associated with H^v\hat{H}_vH^v​ and H^v†\hat{H}_v^\daggerH^v†​ are probed through positive quadratic objectives built from H^v†H^v\hat{H}_v^\dagger \hat{H}_vH^v†​H^v​ and H^vH^v†\hat{H}_v \hat{H}_v^\daggerH^v​H^v†​. The analysis shows that these two searches do not merely produce different approximations to the same variational state. Instead, they select distinct solution sectors. One sector is peaked on flat connections and appears to describe a non-degenerate spatial quantum geometry. Numerical analysis shows this sector to be consistent with a discrete approximation to a Dittrich-Geiller-type vacuum. The other sector consists of states in the kernel of the volume operator. These states are non-flat and exhibit correlations across the graph, consistent with volume-zero excitations over an Ashtekar-Lewandowski-type vacuum. The relation between the two sectors is asymmetric in the sense that while the Ashtekar-Lewandowski-like states are often solutions for both orderings of the quadratic constraints, the Dittrich-Geiller-like states are not. Thus, one can conclude that the ordering of the Hamiltonian constraint affects which sectors are accessible or selected in the variational search, and it suggests that different orderings may exclude some classes of solutions. On the covariant side, this dissertation develops neural network surrogates for spin foam vertex amplitudes, with the aim of approximating the local building blocks of the path-integral dynamics. Starting with the Euclidean Barrett-Crane amplitude, the numerical problem is shown to separate naturally into two physically distinct tasks. The first is to determine which boundary spin configurations are admissible and therefore lie in the support of the amplitude. The second is to approximate the numerical value assigned to each admissible configuration by the vertex amplitude itself. The support is learned with near-perfect accuracy and generalises across cutoffs, showing that the network can infer and extrapolate the known, algorithmically verifiable SU(2)\mathrm{SU}(2)SU(2) recoupling conditions that determine admissibility in the Barrett-Crane case. By contrast, the amplitude values can be reproduced with high precision at fixed cutoff, but do not extrapolate reliably to larger cutoffs. This shows that the dependence of the vertex amplitude on the boundary quantum geometry is considerably more intricate than the structure of its support. The analysis also demonstrates that the representation of discrete spin data is not a merely technical modelling choice but directly affects how well the relevant quantum-geometric information is exposed to the surrogate model. This observation motivates graph-based surrogate architectures in which the spin foam boundary data and their adjacency structure are built into the model more explicitly. These architectures are then applied to the Lorentzian EPRL amplitude whereby in this case, the thesis first addresses the computational bottleneck of exact evaluation by reducing the runtime of the `sl2cfoam-next` exact evaluator sufficiently to generate large training datasets. Preliminary surrogate results show that the support, as well as the sign, of the EPRL amplitude can also be learned effectively, giving first evidence that graph-based learning can capture part of the admissibility structure of the Lorentzian theory.
Abstract We study the near-kernel sector of the Hamiltonian constraint operator in the one-vertex model of quantum reduced loop gravity using exact diagonalisation methods and variational Monte Carlo methods with neural quantum states. The analysis is based on the symmetric Hamiltonian containing both Euclidean and Lorentzian contributions, and on the variational minimization of the positive quadratic operator Q^=C^2\hat{Q}=\hat{C}^2Q^​=C^2, where C^\hat{C}C^ is the symmetric Hamilton constraint, in truncated Hilbert spaces with spin cutoff up to jmax=1001j_{\mathrm{max}}=1001jmax​=1001. In the cases where exact diagonalization is possible, the eigenspace of lowest eigenvalue is one- or two-dimensional, and the eigenvalue, as well as the structure of the eigenstates depends on the cutoff. These ground states are entangled across the three edges. The variational method does not find the lowest eigenspace, but near kernel states, i.e., states with components in a number of low lying eigenspaces. For low cutoffs, they also do not factorize across the three edge degrees of freedom. At larger cutoffs, it increasingly finds near kernel states that factorize over the edges, matched with near-unit fidelity by reduced Thiemann coherent states. We discuss some indication that these states are the optimal near-kernel states if factorization is assumed.
The classical simulation of quantum systems typically requires exponential resources. Recently, the introduction of a machine learning-based wavefunction ansatz has led to the ability to solve the quantum many-body problem in regimes that had previously been intractable for existing exact numerical methods. Here, we demonstrate the utility of the variational representation of quantum states based on artificial neural networks for performing quantum optimization. We show empirically that this methodology achieves high approximation ratio solutions with polynomial classical computing resources for a range of instances of the Maximum Cut (MaxCut) problem whose solutions have been encoded into the ground state of quantum many-body systems up to and including 256 qubits.
Gauge symmetries play a key role in physics appearing in areas such as quantum field theories of the fundamental particles and emergent degrees of freedom in quantum materials. Motivated by the desire to efficiently simulate many-body quantum systems with exact local gauge invariance, gauge equivariant neural-network quantum states are introduced, which exactly satisfy the local Hilbert space constraints necessary for the description of quantum lattice gauge theory with Z_{d} gauge group and non-Abelian Kitaev D(G) models on different geometries. Focusing on the special case of Z_{2} gauge group on a periodically identified square lattice, the equivariant architecture is analytically shown to contain the loop-gas solution as a special case. Gauge equivariant neural-network quantum states are used in combination with variational quantum Monte Carlo to obtain compact descriptions of the ground state wave function for the Z_{2} theory away from the exactly solvable limit, and to demonstrate the confining or deconfining phase transition of the Wilson loop order parameter.
Neural networks have been proposed as efficient numerical wave function Ans\"atze, which can be used to variationally search a wide range of functional forms for ground-state solutions. These neural network methods are also advantageous in that more variational parameters and system degrees of freedom can be easily added. We benchmark the methodology by using neural networks to study several different integrable bosonic quantum systems in one dimension and compare our results to the exact solutions. While testing the scalability of the procedure to systems with many particles, we also introduce using symmetric function inputs to the neural network to enforce exchange symmetries of indistinguishable particles.