{"ID":2851307,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.20771","arxiv_id":"2510.20771","title":"AlphaFlow: Understanding and Improving MeanFlow Models","abstract":"MeanFlow has recently emerged as a powerful framework for few-step generative modeling trained from scratch, but its success is not yet fully understood. In this work, we show that the MeanFlow objective naturally decomposes into two parts: trajectory flow matching and trajectory consistency. Through gradient analysis, we find that these terms are strongly negatively correlated, causing optimization conflict and slow convergence. Motivated by these insights, we introduce $α$-Flow, a broad family of objectives that unifies trajectory flow matching, Shortcut Model, and MeanFlow under one formulation. By adopting a curriculum strategy that smoothly anneals from trajectory flow matching to MeanFlow, $α$-Flow disentangles the conflicting objectives, and achieves better convergence. When trained from scratch on class-conditional ImageNet-1K 256x256 with vanilla DiT backbones, $α$-Flow consistently outperforms MeanFlow across scales and settings. Our largest $α$-Flow-XL/2+ model achieves new state-of-the-art results using vanilla DiT backbones, with FID scores of 2.58 (1-NFE) and 2.15 (2-NFE).","short_abstract":"MeanFlow has recently emerged as a powerful framework for few-step generative modeling trained from scratch, but its success is not yet fully understood. In this work, we show that the MeanFlow objective naturally decomposes into two parts: trajectory flow matching and trajectory consistency. Through gradient analysis,...","url_abs":"https://arxiv.org/abs/2510.20771","url_pdf":"https://arxiv.org/pdf/2510.20771v1","authors":"[\"Huijie Zhang\",\"Aliaksandr Siarohin\",\"Willi Menapace\",\"Michael Vasilkovsky\",\"Sergey Tulyakov\",\"Qing Qu\",\"Ivan Skorokhodov\"]","published":"2025-10-23T17:45:06Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.LG\"]","methods":"[]","has_code":false}