SciPost Physics

SciPost Phys. 10, 147 (2021) · published 17 June 2021

• doi: 10.21468/SciPostPhys.10.6.147
• pdf
• BiBTeX
• RIS
• Submissions/Reports
• 

Abstract

Strongly interacting quantum systems described by non-stoquastic Hamiltonians exhibit rich low-temperature physics. Yet, their study poses a formidable challenge, even for state-of-the-art numerical techniques. Here, we investigate systematically the performance of a class of universal variational wave-functions based on artificial neural networks, by considering the frustrated spin-$1/2$ $J_1-J_2$ Heisenberg model on the square lattice. Focusing on neural network architectures without physics-informed input, we argue in favor of using an ansatz consisting of two decoupled real-valued networks, one for the amplitude and the other for the phase of the variational wavefunction. By introducing concrete mitigation strategies against inherent numerical instabilities in the stochastic reconfiguration algorithm we obtain a variational energy comparable to that reported recently with neural networks that incorporate knowledge about the physical system. Through a detailed analysis of the individual components of the algorithm, we conclude that the rugged nature of the energy landscape constitutes the major obstacle in finding a satisfactory approximation to the ground state wavefunction, and prevents learning the correct sign structure. In particular, we show that in the present setup the neural network expressivity and Monte Carlo sampling are not primary limiting factors.

• Pescia et al., Neural-network quantum states for periodic systems in continuous space
  Phys. Rev. Research 4, 023138 (2022) [Crossref]
• Mansky et al.,
  , 485 (2023) [Crossref]
• Reh et al., Optimizing design choices for neural quantum states
  Phys. Rev. B 107, 195115 (2023) [Crossref]
• Astrakhantsev et al., Broken-Symmetry Ground States of the Heisenberg Model on the Pyrochlore Lattice
  Phys. Rev. X 11, 041021 (2021) [Crossref]
• Zou et al., Learning a compass spin model with neural network quantum states
  J. Phys.: Condens. Matter 34, 125802 (2022) [Crossref]
• Zheng et al.,
  , 236 (2023) [Crossref]
• Viteritti et al., Accuracy of restricted Boltzmann machines for the one-dimensional $J_1-J_2$ Heisenberg model
  SciPost Phys. 12, 166 (2022) [Crossref]
• Zheng et al., Speeding Up Learning Quantum States Through Group Equivariant Convolutional Quantum Ansätze
  PRX Quantum 4, 020327 (2023) [Crossref]
• Ledinauskas et al., Scalable imaginary time evolution with neural network quantum states
  SciPost Phys. 15, 229 (2023) [Crossref]
• Gnech et al., Distilling the Essential Elements of Nuclear Binding via Neural-Network Quantum States
  Phys. Rev. Lett. 133, 142501 (2024) [Crossref]
• Wu et al., From tensor-network quantum states to tensorial recurrent neural networks
  Phys. Rev. Research 5, L032001 (2023) [Crossref]
• Bortone et al., Impact of conditional modelling for a universal autoregressive quantum state
  Quantum 8, 1245 1245 (2024) [Crossref]
• Medvidović et al., Neural-network quantum states for many-body physics
  Eur. Phys. J. Plus 139, 631 (2024) [Crossref]
• Viteritti et al., Transformer Variational Wave Functions for Frustrated Quantum Spin Systems
  Phys. Rev. Lett. 130, 236401 (2023) [Crossref]
• Sauvage et al., Building spatial symmetries into parameterized quantum circuits for faster training
  Quantum Sci. Technol. 9, 015029 (2024) [Crossref]
• Schmitt et al., jVMC: Versatile and performant variational Monte Carlo leveraging automated differentiation and GPU acceleration
  SciPost Phys. Codebases, 2 (2022) [Crossref]
• Moss et al., Enhancing variational Monte Carlo simulations using a programmable quantum simulator
  Phys. Rev. A 109, 032410 (2024) [Crossref]
• Kuang et al., Theory-guided neural network for studying the ground state of 2D spin-orbit coupled Bose–Einstein condensates
  Results in Physics 64, 107935 107935 (2024) [Crossref]
• Vicentini et al., NetKet 3: Machine Learning Toolbox for Many-Body Quantum Systems
  SciPost Phys. Codebases, 7 (2022) [Crossref]
• Inack et al., Neural Annealing and Visualization of Autoregressive Neural Networks in the Newman–Moore Model
  Condensed Matter 7, 38 (2022) [Crossref]
• Roth et al., High-accuracy variational Monte Carlo for frustrated magnets with deep neural networks
  Phys. Rev. B 108, 054410 (2023) [Crossref]
• Klassert et al., Variational learning of quantum ground states on spiking neuromorphic hardware
  iScience 25, 104707 104707 (2022) [Crossref]
• Czischek et al., Data-enhanced variational Monte Carlo simulations for Rydberg atom arrays
  Phys. Rev. B 105, 205108 (2022) [Crossref]
• Andriushchenko et al., A New Look at the Spin Glass Problem from a Deep Learning Perspective
  Entropy 24, 697 (2022) [Crossref]
• Cao et al., Principle of learning sign rules by neural networks in qubit lattice models
  Phys. Rev. B 108, 195127 (2023) [Crossref]
• Yu 于 et al., Solving Quantum Many-Particle Models with Graph Attention Network
  Chinese Phys. Lett. 41, 030202 (2024) [Crossref]
• Bernheimer et al., Determinant- and derivative-free quantum Monte Carlo within the stochastic representation of wavefunctions
  Rep. Prog. Phys. 87, 118001 (2024) [Crossref]
• Klassert et al., Variational Learning of Quantum Ground States on Spiking Neuromorphic Hardware
  SSRN Journal, (2022) [Crossref]
• Chen et al., Empowering deep neural quantum states through efficient optimization
  Nat. Phys. 20, 1476 (2024) [Crossref]
• Rath et al., Quantum Gaussian process state: A kernel-inspired state with quantum support data
  Phys. Rev. Research 4, 023126 (2022) [Crossref]
• Korutcheva et al., The Restricted Boltzmann Machine Ansatz through Adiabiatic Routes
  J. Phys.: Conf. Ser. 2436, 012001 (2023) [Crossref]
• Hibat-Allah et al., Investigating topological order using recurrent neural networks
  Phys. Rev. B 108, 075152 (2023) [Crossref]
• Park et al., Expressive power of complex-valued restricted Boltzmann machines for solving nonstoquastic Hamiltonians
  Phys. Rev. B 106, 134437 (2022) [Crossref]
• Pescia et al., Message-passing neural quantum states for the homogeneous electron gas
  Phys. Rev. B 110, 035108 (2024) [Crossref]
• Zhang et al., Variational Quantum-Neural Hybrid Eigensolver
  Phys. Rev. Lett. 128, 120502 (2022) [Crossref]
• Arnold et al., Replacing Neural Networks by Optimal Analytical Predictors for the Detection of Phase Transitions
  Phys. Rev. X 12, 031044 (2022) [Crossref]
• Inui et al., Determinant-free fermionic wave function using feed-forward neural networks
  Phys. Rev. Research 3, 043126 (2021) [Crossref]
• Hofmann et al., Role of stochastic noise and generalization error in the time propagation of neural-network quantum states
  SciPost Phys. 12, 165 (2022) [Crossref]
• Lovato et al., Hidden-nucleons neural-network quantum states for the nuclear many-body problem
  Phys. Rev. Research 4, 043178 (2022) [Crossref]
• Sarder et al., Feed-forward neural network based variational wave function for the fermionic Hubbard model in one dimension
  J. Phys.: Condens. Matter 34, 375901 (2022) [Crossref]
• Bennewitz et al., Neural Error Mitigation of Near-Term Quantum Simulations
  Nat Mach Intell 4, 618 (2022) [Crossref]
• Lange et al., From architectures to applications: a review of neural quantum states
  Quantum Sci. Technol. 9, 040501 (2024) [Crossref]
• Albash et al., Quantum-inspired tempering for ground state approximation using artificial neural networks
  SciPost Phys. 14, 121 (2023) [Crossref]
• Rzadkowski et al., Artificial neural network states for nonadditive systems
  Phys. Rev. B 106, 155127 (2022) [Crossref]
• Rath et al., Framework for efficient ab initio electronic structure with Gaussian Process States
  Phys. Rev. B 107, 205119 (2023) [Crossref]
• Lange et al., Neural network approach to quasiparticle dispersions in doped antiferromagnets
  Commun Phys 7, 187 (2024) [Crossref]