arXiv:2610.01190
Rapid mixing of quantum spin chains at any finite temperature
Leeseok Kim
Abstract
We prove that a quasi-local quantum Gibbs sampler rapidly mixes to the Gibbs state of any 1D finite-range Hamiltonian at all finite temperatures in $\mathcal{O}\left(\log (n/\varepsilon)\right)$ time. Each update traces out a contiguous block of constant length and reconstructs it using the Petz recovery map, following early proposals for quantum heat-bath Gibbs samplers by Kastoryano and Brandão~\cite{kastoryano2016quantum}. Our proof directly bounds the distance between the updated state and the target Gibbs state using quantum Wasserstein distance of order $1$. We further provide $\mathcal{O}(\text{polylog}(n/\varepsilon))$ depth of quantum circuits to prepare these Gibbs states. We hope these techniques will facilitate the analysis and design of other Gibbs samplers.
AI Summary
The paper asks whether a quantum Gibbs sampler can mix in logarithmic time for general noncommuting one-dimensional spin systems at every fixed finite temperature. It considers bounded-strength, finite-range Hamiltonians with fixed local dimension, without translation-invariance or commutativity assumptions. The sampler updates contiguous blocks by tracing them out and reconstructing them with the Petz recovery map associated with the target Gibbs state. For a sufficiently large block length $\ell_\star$ independent of system size, the resulting quasi-local, KMS-detailed-balanced Lindbladian satisfies $t_{\mathrm{mix}}(\varepsilon)=O(\log(n/\varepsilon))$ in trace distance, with each site updated at total rate one. The proof establishes exponential contraction in quantum Wasserstein distance $W_1$. Updates containing a single-site discrepancy erase it exactly; other updates can increase its Wasserstein cost only by an amount decaying exponentially with distance. Summing this exterior influence in 1D yields contraction rate $\gamma_\ell=1-a_\eta/\ell>0$. The influence estimate follows from boundary localization of the conditional Petz amplitude and boundary-to-interior factorization of Gibbs states, using complex interpolation, Lieb–Robinson bounds, and existing 1D Gibbs-state estimates. An implementation using logarithmic-size Hamiltonian windows yields randomized nearest-neighbor circuits with $O(n\,\mathrm{polylog}(n/\varepsilon))$ gates and polylogarithmic depth. Temperature and local dimension are fixed: their resource dependence is not quantified sharply, and the stated implementation overhead appears at least triply exponential in inverse temperature. Extensions to other samplers, long-range interactions, and higher dimensions remain open. A note added reports concurrent constant-depth Gibbs-state preparation, improving the circuit-depth bound but addressing a different question from sampler mixing.
In particular, it helped us resolve the issue via the unitalization argument in Section~\ref{subsubsec:unitalization}, prove Assumption~\ref{asm:QL} in Section~\ref{sec:verify-QL}, and improve the algorithm design and analysis in Appendix~\ref{app:implementation}.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.