Neural-network-assisted Monte Carlo sampling trained by Quantum Approximate Optimization Algorithm

Abstract

Sampling problems are widely regarded as the task for which quantum computers can most readily provide a quantum advantage. Leveraging this feature, the quantum-enhanced Markov chain Monte Carlo [Layden, D. et al., Nature 619, 282-287 (2023)] has been proposed recently, where sampling from a quantum computer is used as a proposal distribution and convergence to a target distribution is accelerated. However, guaranteeing convergence to the target distribution typically forces one to impose restrictive symmetry constraints on the quantum circuit, which makes it hard to design good proposal distributions and prevents making full use of the advantage of a quantum computer. We explore a hybrid quantum-classical MCMC framework that combines a quantum circuit with a generative neural sampler (GNS). The GNS is trained on quantum samples and acts as a classical surrogate to efficiently emulate quantum outputs, thereby lifting circuit constraints. We apply this method to Boltzmann sampling of spin glasses using proposals trained with a QAOA circuit. This approach outperforms conventional methods, showing a $\sim$100$\times$ improvement in spectral gap over uniform proposals. Notably, it maintains similar acceleration even without parameter optimization. These results establish the method as a viable sampling-based quantum algorithm for NISQ devices and highlight its potential for solving practical problems with quantum computation.

12 pages, 8 figures

References in corpus (34)

• SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python
• Quantum Computing in the NISQ era and beyond
• Supplementary information for "Quantum supremacy using a programmable superconducting processor"
• A variational eigenvalue solver on a quantum processor
• Quantum algorithm for solving linear systems of equations
• Variational Quantum Algorithms
• Quantum computational advantage using photons
• Barren plateaus in quantum neural network training landscapes
• Quantum Circuit Learning
• Qulacs: a fast and versatile quantum circuit simulator for research purpose
• Quantum critical dynamics in a 5000-qubit programmable spin glass
• Computational advantage of quantum random sampling
• A Cluster Monte Carlo Algorithm for 2-Dimensional Spin Glasses
• Measurements as a roadblock to near-term practical quantum advantage in chemistry: resource analysis
• Asymptotically unbiased estimation of physical observables with neural samplers
• Efficient Cluster Algorithm for Spin Glasses in Any Space Dimension
• Parameter Transfer for Quantum Approximate Optimization of Weighted MaxCut
• Chemistry Beyond the Scale of Exact Diagonalization on a Quantum-Centric Supercomputer
• Quantum-enhanced Markov chain Monte Carlo
• Boosting Monte Carlo simulations of spin glasses using autoregressive neural networks
• Effective quantum volume, fidelity and computational cost of noisy quantum processing experiments
• Scaling overhead of embedding optimization problems in quantum annealing
• Thermalization, freeze-out and noise: deciphering experimental quantum annealers
• Programmable Quantum Annealers as Noisy Gibbs Samplers
• Quantum Approximate Optimization Algorithm pseudo-Boltzmann states
• Machine-learning-assisted Monte Carlo fails at sampling computationally hard problems
• Sampling, rates, and reaction currents through reverse stochastic quantization on quantum computers
• Trainability Barriers in Low-Depth QAOA Landscapes
• QAOA-MC: Markov chain Monte Carlo enhanced by Quantum Alternating Operator Ansatz
• Accelerating equilibrium spin-glass simulations using quantum annealers via generative deep learning
• The autoregressive neural network architecture of the Boltzmann distribution of pairwise interacting spins systems
• Exploring the neighborhood of 1-layer QAOA with Instantaneous Quantum Polynomial circuits
• From quantum-enhanced to quantum-inspired Monte Carlo
• Self-learning Monte Carlo with equivariant Transformer