Classifier-Free Guidance Is a Predictor-Corrector Apple Machine Learning Research
[[{“value”:”This paper was accepted at the Mathematics of Modern Machine Learning (M3L) Workshop at NeurIPS 2024. We investigate the unreasonable effectiveness 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… Read More »Classifier-Free Guidance Is a Predictor-Corrector Apple Machine Learning Research