{"id":"arxiv/2609.08873","url":"https://arxiv.org/abs/2609.08873","tags":["cs.DS","cs.LG","math.PR","math.ST","stat.ML"],"score":0,"title":"High-Magnetization Sampling at Low Temperatures: Ising Models and Bayesian Sparse Linear Regression","domain":"cs.DS","source":"arxiv","authors":["Syamantak Kumar","Purnamrita Sarkar","Kevin Tian","Yusong Zhu"],"reviews":0,"subject":"cs.DS","summary":"Sparsity is a powerful structural resource in optimization and statistics. We develop frameworks for leveraging sparsity in sampling problems over the Hamming slice $\\mathcal{X}_k^d:=\\{\\mathbf{x}\\in\\{\\pm 1\\}^d:|\\{i:\\mathbf{x}_i=1\\}|=k\\}$, in high-dimensional regimes where $k\\ll d$ (i.e., where $\\mathcal{X}_k^d$ is \\emph{highly magnetized}). We use our frameworks to design improved samplers for canonical problems in the study of \\emph{Ising models} and \\emph{Bayesian sparse linear regression}. Our first main result considers the \\emph{Sherrington--Kirkpatrick} (SK) model restricted to…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-09-08","reactions":{}}