Apple's Normalizing Trajectory Models hit top image quality in four steps

Diffusion-based models typically decompose sampling into many small Gaussian denoising steps, an assumption that breaks down when generation is compressed to a few coarse transitions. Existing few-step methods address this through distillation, consistency training, or adversarial objectives, but sacrifice the likelihood framework in the process. NTM models each reverse step as an expressive conditional normalizing flow with exact likelihood training.

Architecturally, NTM combines shallow invertible blocks within each step with a deep parallel predictor across the trajectory, forming an end-to-end network. Its exact trajectory likelihood enables self-distillation: a lightweight denoiser trained on the score function induced by the model itself produces high-quality samples in four steps.