High-Accuracy Sampling for Diffusion Models and Log-Concave Distributions

Unbiased stochastic queries can support sampling whose cost grows only polylogarithmically with inverse target error. This talk applies the framework to diffusion generation with sufficiently accurate score estimates and to log-concave sampling with subexponential gradient noise. The latter result contrasts with stochastic convex optimization, where even simple additive noise can require polynomially many queries.