about
Step-by-Step Diffusion: An Elementary Tutorial (arxiv.org)
2 points by Anon84 305 days ago | hide | past | pdf | discuss on HN

In plain words: A beginner-friendly course that teaches how AI turns random noise into images or text step by step, building the needed math from scratch. Unlike typical introductions that dive into heavy equations, it simplifies the details yet still derives algorithms that work correctly.

Abstract

We present an accessible first course on diffusion models and flow matching for machine learning, aimed at a technical audience with no diffusion experience. We try to simplify the mathematical details as much as possible (sometimes heuristically), while retaining enough precision to derive correct algorithms.

Preetum Nakkiran, Arwen Bradley, Hattie Zhou, Madhu Advani
arXiv:2406.08929 · cs.LG, cs.AI, cs.CV, stat.ML · submitted Jun 13, 2024 · updated Jun 23, 2024
abstract · pdf · html · 35 pages, 11 figures

add comment on HN
Also discussed: Jul 2024 (2 points, 0 comments) · Jun 2024 (3 points, 0 comments) · Jun 2024 (3 points, 0 comments)