{"id":"arxiv/2603.03159","url":"https://arxiv.org/abs/2603.03159","tags":["hep-lat","hep-th"],"score":0,"title":"Solving sign problems with physics-informed kernels","domain":"hep-lat","source":"arxiv","authors":["Friederike Ihssen","Renzo Kapust","Jan M. Pawlowski"],"reviews":0,"subject":"hep-lat","summary":"In the present work we construct a novel generative architecture for systems with complex probability distributions. In general, these sampling tasks come with two challenges: resolving sign problems and efficient sampling. The architecture is based on physics-informed kernels (PIKs) introduced in arXiv:2510.26678, and aims at resolving both challenges. Key to the complex PIK-architecture is its probability-weight preserving property, which allows us to map the sampling task to one on a sign-problem free manifold with a simple distribution and efficient sampling. The potential of this novel…","category":"research","comments":0,"disputed":false,"division":"physics","mentions":0,"discussed":0,"published":"2026-03-03","reactions":{}}