AI RESEARCH
A Unified Measure-Theoretic View of Diffusion, Score-Based, and Flow Matching Generative Models
arXiv CS.LG
•
ArXi:2605.06829v1 Announce Type: new We survey continuous-time generative modeling methods based on transporting a simple reference distribution to a data distribution via stochastic or deterministic dynamics. We present a unified framework in which diffusion models, score-based generative models, and flow matching are instances of learning a time-dependent vector field that induces a family of marginals $(\rho_t)_{t \in [0,1]}$ governed by continuity and Fokker-Planck equations.