AI RESEARCH

From Score Matching to Diffusion: A Fine-Grained Error Analysis in the Gaussian Setting

arXiv CS.LG

ArXi:2503.11615v3 Announce Type: replace Sampling from an unknown distribution, accessible only through discrete samples, is a fundamental problem at the core of generative AI. The current state-of-the-art methods follow a two-step process: first, estimating the score function (the gradient of a smoothed log-distribution) and then applying a diffusion-based sampling algorithm -- such as Langevin or Diffusion models.