Concept
Classifier-free guidance
Classifier-free guidance (CFG) is a sampling trick for conditional generative models (diffusion, flow matching) that sharpens how strongly the output follows its conditioning, such as a text prompt, without training a separate classifier.
The rest of “Classifier-free guidance” is a premium feature: every concept in the library gets a precise, practitioner-focused write-up like this one, cross-linked straight from the paper summaries that use it.
Log in to unlock→