Sampling with Shielded Langevin Monte Carlo Using Navigation Potentials

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Zilberstein, Nicolas, Segarra, Santiago, Chamon, Luiz
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912790796566528
author Zilberstein, Nicolas
Segarra, Santiago
Chamon, Luiz
author_facet Zilberstein, Nicolas
Segarra, Santiago
Chamon, Luiz
contents We introduce shielded Langevin Monte Carlo (LMC), a constrained sampler inspired by navigation functions, capable of sampling from unnormalized target distributions defined over punctured supports. In other words, this approach samples from non-convex spaces defined as convex sets with convex holes. This defines a novel and challenging problem in constrained sampling. To do so, the sampler incorporates a combination of a spatially adaptive temperature and a repulsive drift to ensure that samples remain within the feasible region. Experiments on a 2D Gaussian mixture and multiple-input multiple-output (MIMO) symbol detection showcase the advantages of the proposed shielded LMC in contrast to unconstrained cases.
format Preprint
id arxiv_https___arxiv_org_abs_2512_22153
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Sampling with Shielded Langevin Monte Carlo Using Navigation Potentials
Zilberstein, Nicolas
Segarra, Santiago
Chamon, Luiz
Computation
Machine Learning
We introduce shielded Langevin Monte Carlo (LMC), a constrained sampler inspired by navigation functions, capable of sampling from unnormalized target distributions defined over punctured supports. In other words, this approach samples from non-convex spaces defined as convex sets with convex holes. This defines a novel and challenging problem in constrained sampling. To do so, the sampler incorporates a combination of a spatially adaptive temperature and a repulsive drift to ensure that samples remain within the feasible region. Experiments on a 2D Gaussian mixture and multiple-input multiple-output (MIMO) symbol detection showcase the advantages of the proposed shielded LMC in contrast to unconstrained cases.
title Sampling with Shielded Langevin Monte Carlo Using Navigation Potentials
topic Computation
Machine Learning
url https://arxiv.org/abs/2512.22153