AI RESEARCH

Large Scale Diffusion Distillation via Score-Regularized Continuous-Time Consistency

arXiv CS.LG

ArXi:2510.08431v3 Announce Type: replace-cross Although continuous-time consistency models (e.g., sCM, MeanFlow) are theoretically principled and empirically powerful for fast academic-scale diffusion, its applicability to large-scale text-to-image and video tasks remains unclear due to infrastructure challenges in Jacobian-vector product (JVP) computation and the limitations of evaluation benchmarks like