Reproduce “Quantum-enhanced Markov chain Monte Carlo” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
David Layden, Guglielmo Mazzola, Ryan V. Mishmash, Mario Motta, Pawel Wocjan, Jin-Sung Kim, Sarah Sheldon
IBM Quantum researchers developed a hybrid quantum-classical Markov Chain Monte Carlo (MCMC) algorithm to accelerate sampling from complex classical Ising model distributions, particularly in low-temperature spin-glass regimes. The method, experimentally demonstrated on superconducting quantum processors, achieved substantially better convergence rates and faster estimation of thermal averages compared to classical local and uniform MCMC strategies.
Key insights
- The quantum algorithm generates "best of both worlds" proposals, characterized by small absolute energy differences (leading to high acceptance rates) and large Hamming distances (enabling efficient exploration of rugged energy landscapes).
- Simulations revealed a polynomial enhancement in the average absolute spectral gap (δ), a key metric for convergence, indicating a potential scaling advantage over classical MCMC methods for larger problem sizes.
- Experimental implementation on noisy intermediate-scale quantum (NISQ) hardware for up to 10 qubits demonstrated a concrete quantum enhancement in MCMC convergence, showing that quantum devices can improve useful computational tasks even with current limitations.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.