{"id":"arxiv/2602.01449","url":"https://arxiv.org/abs/2602.01449","tags":["math.NA","math.PR","stat.ML"],"score":0,"title":"Dimension-Free Multimodal Sampling via Preconditioned Annealed Langevin Dynamics","domain":"math.NA","source":"arxiv","authors":["Lorenzo Baldassari","Josselin Garnier","Knut Solna","Maarten V. de Hoop"],"reviews":0,"subject":"math.NA","summary":"Designing sampling algorithms for multimodal targets that remain stable under refinement of the finite-dimensional approximation of an underlying function-space problem is a central challenge. Annealed Langevin dynamics (ALD) is a natural alternative to classical Langevin in this context, since it is often observed to improve exploration across modes. Yet a gap remains between its empirical success and existing theory: under which conditions can ALD be guaranteed to remain stable across dimensions? In this paper, we bridge this gap by providing a uniform-in-dimension analysis of…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-02-01","reactions":{}}