arXiv Open Access 2025

Contractive kinetic Langevin samplers beyond global Lipschitz continuity

Iosif Lytras Panayotis Mertikopoulos

Lihat Sumber

Abstrak

In this paper, we examine the problem of sampling from log-concave distributions with (possibly) superlinear gradient growth under kinetic (underdamped) Langevin algorithms. Using a carefully tailored taming scheme, we propose two novel discretizations of the kinetic Langevin SDE, and we show that they are both contractive and satisfy a log-Sobolev inequality. Building on this, we establish a series of non-asymptotic bounds in $2$-Wasserstein distance between the law reached by each algorithm and the underlying target measure.

Topik & Kata Kunci

math.PR math.NA stat.ML

Penulis (2)

Iosif Lytras

Panayotis Mertikopoulos

Format Sitasi

APA MLA BibTeX

Lytras, I., Mertikopoulos, P. (2025). Contractive kinetic Langevin samplers beyond global Lipschitz continuity. https://arxiv.org/abs/2509.12031

Akses Cepat

Lihat di Sumber

Informasi Jurnal

Tahun Terbit: 2025
Bahasa: en
Sumber Database: arXiv
Akses: Open Access ✓