Metraipolis › archive › dc_78dff60b662e

MEND: RL For Flow Models via Proximal Velocity Matching — arXiv:2610.05954

Source
arXiv
Original
https://arxiv.org/abs/2610.05954
License (as archived)
CC0-1.0
Archived
2026-10-06T05:31:22.250Z
SHA-256 of the archived text
71c6af4a07e6bea9df8c383779acc39311d292e09c3c6a5ebde278b598df0005
Stable URL of this page
https://metraipolis.com/documento/dc_78dff60b662e

How to cite

MEND: RL For Flow Models via Proximal Velocity Matching — arXiv:2610.05954 — arXiv, https://arxiv.org/abs/2610.05954 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_78dff60b662e (Metraipolis archive dc_78dff60b662e, sha256 71c6af4a07e6bea9df8c383779acc39311d292e09c3c6a5ebde278b598df0005)
BibTeX
@misc{dc78dff60b662e,
  title = {MEND: RL For Flow Models via Proximal Velocity Matching — arXiv:2610.05954},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.05954},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_78dff60b662e; sha256 71c6af4a07e6bea9df8c383779acc39311d292e09c3c6a5ebde278b598df0005},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.05954. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Reward post-training of flow models either reweights the model's own samples under a KL penalty or a frozen reference, often for thousands of updates, or backpropagates the reward and moves every sample without checking that the move is worth its size. We introduce MEND, a reinforcement learning method built on proximal velocity matching. MEND caps rewards within each prompt group, so samples that already score well receive no move. Below the cap, it proposes moves along the reward gradient and accepts one only when its capped reward gain exceeds a quadratic displacement price. The model then regresses onto the resulting velocity targets, with no KL term, frozen reference model, or advantage weights. In 100 updates, MEND outperforms Flow-GRPO (about 4k updates) on five of six evaluators at the same distance to base-model images. Under an equal-budget protocol, it surpasses ReFL and DiffusionNFT at every evaluated update across four training rewards, reaching PickScore 24.03 versus 23.92 and 23.43, respectively. A 300-update three-reward run also surpasses the five-reward DiffusionNFT model on all three rewards it trains on. MEND is general and easy to adopt: it applies to any flow backbone with a differentiable reward.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.05954 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-06T05:27:05.843Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).