Classifier-Free Guidance is a Predictor-Corrector

Arwen Bradley, Preetum Nakkiran

Keywords: diffusion, guidance, theory, SDE, Langevin, score

TL;DR: We prove that classifier-free guidance is equivalent to interlacing a diffusion denoiser with Langevin dynamics, thus giving a more principled interpretation to CFG.

Abstract: We investigate the theoretical foundations of classifier-free guidance (CFG). CFG is the dominant method of conditional sampling for text-to-image diffusion models, yet unlike other aspects of diffusion, it remains on shaky theoretical footing. In this paper, we disprove common misconceptions, by showing that CFG interacts differently with DDPM and DDIM, and neither sampler with CFG generates the gamma-powered distribution $p(x|c)^\gamma p(x)^{1-\gamma}$. Then, we clarify the behavior of CFG by showing that it is a kind of predictor-corrector method (Song et al., 2020) that alternates between denoising and sharpening, which we call predictor-corrector guidance (PCG). We prove that in the SDE limit, CFG is actually equivalent to combining a DDIM predictor for the conditional distribution together with a Langevin dynamics corrector for a gamma-powered distribution (with a carefully chosen gamma). Our work thus provides a lens to theoretically understand CFG by embedding it in a broader design space of principled sampling methods.

Is Neurips Submission: No

Submission Number: 5