Apple's Normalizing Trajectory Models generate images in four steps with exact likelihood
Apple Machine Learning Research published Normalizing Trajectory Models (NTM) at NeurIPS, which models each reverse diffusion step as an expressive conditional normalizing flow with exact likelihood training, pairing shallow invertible blocks per step with a deep parallel predictor across the trajectory and trainable from scratch or from pretrained flow-matching models. Its exact trajectory likelihood also enables self-distillation: a lightweight denoiser trained on the score function induced by the model itself produces high-quality samples in four steps. On text-to-image benchmarks, NTM matches or outperforms strong image generation baselines in just four sampling steps while uniquely retaining exact likelihood over the generative trajectory.
Few-step generation methods usually abandon the likelihood framework; NTM suggests exact likelihood and few-step, high-quality sampling can coexist.
content type paperpublished October 2026
Normalizing Trajectory Models
AuthorsJiatao Gu†, Tianrong Chen, Ying Shen‡**, David Berthelot, Shuangfei Zhai, Josh Susskind
Diffusion-based models 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. We introduce Normalizing Trajectory Models (NTM), which 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 trainable from scratch or initializable from pretrained flow-matching models. Its exact trajectory likelihood further enables self-distillation: a lightweight denoiser trained on the score function induced by the model itself produces high-quality samples in four steps. On text-to-image benchmarks, NTM matches or outperforms strong image generation baselines in just four sampling steps while uniquely retaining exact likelihood over the generative trajectory.
- † University of Pennsylvania
- ‡ UIUC
- ** Work done while at Apple