Scalable Generative Sampling and Multilevel Estimation for Lattice Field Theories Near Criticality

A. Singha, J. Kauffmann, E. Cellini, K. Jansen, S. Nakajima · 2026

Read Scalable Generative Sampling and Multilevel Estimation for Lattice Field Theories Near Criticality by A. Singha, J. Kauffmann, E. Cellini, K. Jansen, S. Nakajima (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-10209.txt.

Description

Singha and colleagues sample lattice field theories near criticality with a multiscale generative sampler inspired by renormalization. Their sampler models the Boltzmann distribution through a coarse-to-fine hierarchy. At each level, a Gaussian mixture captures coarse-field dependence, while a normalizing flow refines the rest. Coarse levels encode long modes and finer levels add short-distance fluctuations. Preserving coarse fields exactly yields restriction maps and unbiased multilevel variance reduction. On phi-fourth theory, autocorrelation times run orders below hybrid Monte Carlo, with unbiased observables matching long hybrid runs.

Catalog record

CREATORA. Singha, J. Kauffmann, E. Cellini, K. Jansen, S. Nakajima
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE65220
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.10209

Browse the full library · About the Tower of Babel