Mean Flows for One-step Generative Modeling

NeurIPSOral2025

Authors
Zhengyang Geng, Mingyang Deng, Xingjian Bai, J Zico Kolter, Kaiming He
Affiliation
Carnegie Mellon University
Venue
NeurIPS 2025
Track
Oral

TL;DR

We propose a principled and effective framework for one-step generative modeling. We introduce the notion of average velocity to characterize flow fields, in contrast to instantaneous velocity modeled by Flow Matching methods.

Opening excerpt from the authors’ abstract. source

Read the paper

Topics

generative model flow matching rag

← All NeurIPS 2025 Oral papers · Browse the whole archive