{"ID":2863859,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.25170","arxiv_id":"2509.25170","title":"GLASS Flows: Transition Sampling for Alignment of Flow and Diffusion Models","abstract":"The performance of flow matching and diffusion models can be greatly improved at inference time using reward alignment algorithms, yet efficiency remains a major limitation. While several algorithms were proposed, we demonstrate that a common bottleneck is the sampling method these algorithms rely on: many algorithms require to sample Markov transitions via SDE sampling, which is significantly less efficient and often less performant than ODE sampling. To remove this bottleneck, we introduce GLASS Flows, a new sampling paradigm that simulates a \"flow matching model within a flow matching model\" to sample Markov transitions. As we show in this work, this \"inner\" flow matching model can be retrieved from a pre-trained model without any re-training, combining the efficiency of ODEs with the stochastic evolution of SDEs. On large-scale text-to-image models, we show that GLASS Flows eliminate the trade-off between stochastic evolution and efficiency. Combined with Feynman-Kac Steering, GLASS Flows improve state-of-the-art performance in text-to-image generation, making it a simple, drop-in solution for inference-time scaling of flow and diffusion models.","short_abstract":"The performance of flow matching and diffusion models can be greatly improved at inference time using reward alignment algorithms, yet efficiency remains a major limitation. While several algorithms were proposed, we demonstrate that a common bottleneck is the sampling method these algorithms rely on: many algorithms r...","url_abs":"https://arxiv.org/abs/2509.25170","url_pdf":"https://arxiv.org/pdf/2509.25170v3","authors":"[\"Peter Holderrieth\",\"Uriel Singer\",\"Tommi Jaakkola\",\"Ricky T. Q. Chen\",\"Yaron Lipman\",\"Brian Karrer\"]","published":"2025-09-29T17:58:36Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.AI\",\"stat.ML\"]","methods":"[\"Diffusion Model\"]","has_code":false}