Metraipolis › archive › dc_413e4e532153
Learning Noise Optimization for Reward Alignment in Few-Step Generative Models — arXiv:2611.06255
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.06255
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-05T18:22:13.001Z
- SHA-256 of the archived text
9c21ed8ae4a08386df770a7833b716f89dce4021e58c77c3c26185c74ea4b46c- Stable URL of this page
https://metraipolis.com/documento/dc_413e4e532153
How to cite
Learning Noise Optimization for Reward Alignment in Few-Step Generative Models — arXiv:2611.06255 — arXiv (en), https://arxiv.org/abs/2611.06255 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_413e4e532153 (Metraipolis archive dc_413e4e532153, sha256 9c21ed8ae4a08386df770a7833b716f89dce4021e58c77c3c26185c74ea4b46c)
BibTeX
@misc{dc413e4e532153,
title = {Learning Noise Optimization for Reward Alignment in Few-Step Generative Models — arXiv:2611.06255},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.06255},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_413e4e532153; sha256 9c21ed8ae4a08386df770a7833b716f89dce4021e58c77c3c26185c74ea4b46c},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.06255. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Reward alignment for diffusion- and flow-based generative models is often formulated over stochastic trajectories, making it less suitable for few-step deterministic generators. Initial noise optimization provides a trajectory-free alternative, but existing gradient-free methods require costly per-sample search at inference. We introduce LeNO (Learning Noise Optimization), which amortizes reward-guided initial noise optimization into a learned residual transport while keeping the pretrained generator frozen. LeNO learns an anchor-conditioned distribution over noise corrections using Local GRPO and scalar rewards, eliminating iterative test-time optimization. For Gaussian initial noise with norm-preserving projection, we derive a tractable upper bound on the KL divergence between the corrected noise distribution and the original prior, providing a principled regularization objective. We evaluate LeNO on few-step text-to-image and protein structure generation, including expensive black-box designability rewards. Across both domains, LeNO improves target rewards with a single noise correction at inference while largely preserving generation quality and diversity. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2611.06255 — 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 2026-11-05T18:15:49.014Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).