Metraipolis › archive › dc_10239d0ce3d5

Generator-Aligned Reward Modeling with Target-Aware Bounded Optimization for Visual Generation — arXiv:2611.07648

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.07648
License (as archived)
CC0-1.0
Archived
2026-11-06T14:53:27.872Z
SHA-256 of the archived text
6745b177d3e007ee83672f6e71dfd9207538e24a61ac281f8783c4d72c54ba3e
Stable URL of this page
https://metraipolis.com/documento/dc_10239d0ce3d5

How to cite

Generator-Aligned Reward Modeling with Target-Aware Bounded Optimization for Visual Generation — arXiv:2611.07648 — arXiv (en), https://arxiv.org/abs/2611.07648 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_10239d0ce3d5 (Metraipolis archive dc_10239d0ce3d5, sha256 6745b177d3e007ee83672f6e71dfd9207538e24a61ac281f8783c4d72c54ba3e)
BibTeX
@misc{dc10239d0ce3d5,
  title = {Generator-Aligned Reward Modeling with Target-Aware Bounded Optimization for Visual Generation — arXiv:2611.07648},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.07648},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_10239d0ce3d5; sha256 6745b177d3e007ee83672f6e71dfd9207538e24a61ac281f8783c4d72c54ba3e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.07648. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Post-training aims to enhance desired attributes in visual generative models while preserving their general generation capabilities. Reward-guided post-training faces two challenges: real training data may mismatch the generator's output distribution, and attribute supervision may be confounded by unrelated content. Unbounded reward maximization can also cause overoptimization and degrade generation quality. We propose a framework combining Generator-Aligned Reward Modeling (GARM) and Target-Aware Bounded Optimization (TABO). GARM mitigates distribution mismatch through generator-support matching and reduces content confounding using content-matched real contrasts, yielding generator-relevant, attribute-focused reward supervision. TABO maps the learned reward to a smooth, upper-bounded function calibrated from real target-positive examples. This transformation is independent of the optimization algorithm and supports both direct reward-gradient and policy-based optimization. We further introduce Target-Aware GRPO (TA-GRPO), an adaptation of Flow-GRPO that combines fixed stage-wise advantage scaling, a generator-to-target curriculum, and timestep-balanced credit assignment for flow-based video optimization. Experiments show that GARM improves target-attribute expression while maintaining competitive generation quality. TABO better balances attribute enhancement and visual quality than unconstrained reward maximization, while TA-GRPO improves target enhancement over vanilla Flow-GRPO and reduces unintended attribute leakage on neutral prompts. These results highlight the importance of generator-aligned reward supervision and controlled, data-calibrated optimization for targeted visual generation.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2611.07648 — 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 2026-11-06T14:25:42.613Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).