Parts | Days | Selection | Search | Updates | Downloads | Help
SOE: Fachverband Physik sozio-ökonomischer Systeme
SOE 14: Focus Session: Physics of AI I (joint session SOE/DY)
SOE 14.2: Talk
Thursday, March 12, 2026, 10:00–10:15, GÖR/0226
Statistical Physics of Classifier-free Diffusion Guidance — •Enrico Ventura1, Beatrice Achilli1, Carlo Lucibello1, and Luca Ambrogioni2 — 1Bocconi University, Milan, Italy — 2Radboud University, Nijmegen, The Netherlands
Classifier-free Guidance (CFG) is a simple yet effective technique that helps diffusion models better follow a user's prompt. By combining standard unconditional diffusion with diffusion conditioned on a specific class of the data, it steers generation toward samples (e.g. images, videos or text) that more clearly reflect the intended content. We propose a description of the sampling dynamics of a diffusion model under CFG based on the statistical mechanics of disordered systems. Specifically, we study the time-dependent transformation of the diffusion potential providing a quantitative prediction of the way a complex target distribution is deformed to improve data generation. Moreover, we leverage our results to propose alternative theory-based guidance schedules that enhance such beneficial effects.
Keywords: Diffusion Models; Random Energy Model; Disordered Systems; Stochastic Processes; Mean Field Analysis