{"id":"arxiv/2607.24324","url":"https://arxiv.org/abs/2607.24324","tags":["stat.ME","math.ST"],"score":0,"title":"Diffusion Bootstrap for High-Dimensional Linear Models","domain":"stat.ME","source":"arxiv","authors":["Ce Liang","Wei Ma"],"reviews":0,"subject":"stat.ME","summary":"Classical bootstrap methods can behave poorly in high-dimensional linear models: the pairs bootstrap often yields overly conservative inference, whereas the residual bootstrap can be anti-conservative, reflecting systematic failures in variance calibration. We propose a diffusion-based pairs bootstrap that replaces the empirical joint distribution with a learned generative law. We establish variance consistency under a score approximation assumption, using complementary SDE and PDE arguments. Counterexamples show that terminal $W_4$ convergence alone is insufficient for variance consistency.…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-07-27","reactions":{}}