From Navigation to Refinement: Revealing the Two-Stage Nature of Flow-based Diffusion Models through Oracle Velocity

arXiv:2512.02826v3 Announce Type: replace-cross Abstract: Flow-based diffusion models have emerged as a leading paradigm for training generative models across images and videos. However, their memorization-generalization behavior remains poorly understood. In this work, we revisit the flow matching (FM) objective and study its marginal velocity field, which admits a closed-form expression, allowing exact computation of the oracle FM target. Analyzing this oracle velocity field reveals that flow-based diffusion models inherently formulate a two-stage training target: an early stage guided by a mixture of data modes, and a later stage dominated by the nearest data sample. The two-stage objective leads to distinct learning behaviors: the early navigation stage generalizes across data modes to form global layouts, whereas the later refinement stage increasingly memorizes fine-grained details. Leveraging these insights, we explain the effectiveness of practical techniques such as timestep-shifted schedules, classifier-free guidance intervals, and latent space design choices. Our study deepens the understanding of diffusion model training dynamics and offers principles for guiding future architectural and algorithmic improvements. Our project page is available at: https://maps-research.github.io/from-navigation-to-refinement/.

Related

• Less is More: Data-Efficient Adaptation for Controllable Text-to-Video Generation
• ChopGrad: Pixel-Wise Losses for Latent Video Diffusion via Truncated Backpropagation
• DeCo: Frequency-Decoupled Pixel Diffusion for End-to-End Image Generation
• On the Robustness of Diffusion-Based Image Compression to Bit-Flip Errors
• Generative Phomosaic with Structure-Aligned and Personalized Diffusion
• Less is More: Data-Efficient Adaptation for Controllable Text-to-Video Generation

Source: arXiv cs.AI | 2026-04-13