[2603.02984] Variance reduction in lattice QCD observables via normalizing flows

High Energy Physics - Lattice arXiv:2603.02984 (hep-lat) [Submitted on 3 Mar 2026] Title:Variance reduction in lattice QCD observables via normalizing flows Authors:Ryan Abbott, Denis Boyda, Yang Fu, Daniel C. Hackett, Gurtej Kanwar, Fernando Romero-López, Phiala E. Shanahan, Julian M. Urban View a PDF of the paper titled Variance reduction in lattice QCD observables via normalizing flows, by Ryan Abbott and 6 other authors View PDF HTML (experimental) Abstract:Normalizing flows can be used to construct unbiased, reduced-variance estimators for lattice field theory observables that are defined by a derivative with respect to action parameters. This work implements the approach for observables involving gluonic operator insertions in the SU(3) Yang-Mills theory and two-flavor Quantum Chromodynamics (QCD) in four space-time dimensions. Variance reduction by factors of $10$-$60$ is achieved in glueball correlation functions and in gluonic matrix elements related to hadron structure, with demonstrated computational advantages. The observed variance reduction is found to be approximately independent of the lattice volume, so that volume transfer can be utilized to minimize training costs. Comments: Subjects: High Energy Physics - Lattice (hep-lat); Machine Learning (cs.LG) Report number: FERMILAB-PUB-26-0130-T, MIT-CTP/6010 Cite as: arXiv:2603.02984 [hep-lat]   (or arXiv:2603.02984v1 [hep-lat] for this version)   https://doi.org/10.48550/arXiv.2603.02984 Focus to learn more arX...