SOURCE-LINKED INTELLIGENCE

Mean Velocity Matching: Rethinking Generative Dynamics in Diffusion Models

arXiv · AI, language, vision and robotics · article · Sep 21, 2026 · UTC

This work studies prediction parameterization for stochastic generative dynamics in diffusion models. Existing velocity-based generative models provide the simplicity of learning a single transport field, but their standard formulation is deterministic, whereas stochastic extensions generally require additional score information or an intermediate velocity-to-score reconstruction. To retain single-field prediction while directly supporting stochastic reverse dynamics, this paper introduces Mean Velocity Matching (MVM). MVM constructs a Gaussian perturbation process for which the conditional ex

Read original source ↗ Open in workspace

recordType

paper

region

Global