Metraipolis › archive › dc_119f5ec313be

Beyond Random Couplings: Contrastive Noise Alignment in Generative Flows — arXiv:2609.18488

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.18488
License (as archived)
CC0-1.0
Archived
2026-10-15T15:08:01.959Z
SHA-256 of the archived text
c91e10fb8e29582565f248eefe3e55a44c2e8b3306938fa4456b7cc99af658e2
Stable URL of this page
https://metraipolis.com/documento/dc_119f5ec313be

How to cite

Beyond Random Couplings: Contrastive Noise Alignment in Generative Flows — arXiv:2609.18488 — arXiv (en), https://arxiv.org/abs/2609.18488 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_119f5ec313be (Metraipolis archive dc_119f5ec313be, sha256 c91e10fb8e29582565f248eefe3e55a44c2e8b3306938fa4456b7cc99af658e2)
BibTeX
@misc{dc119f5ec313be,
  title = {Beyond Random Couplings: Contrastive Noise Alignment in Generative Flows — arXiv:2609.18488},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.18488},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_119f5ec313be; sha256 c91e10fb8e29582565f248eefe3e55a44c2e8b3306938fa4456b7cc99af658e2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.18488. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Diffusion and flow-matching models are typically trained by corrupting data through independently sampled Gaussian noise. While simple and scalable, this forward process induces arbitrary data-noise couplings, forcing the network to learn high-curvature transports between unrelated endpoints. Existing optimal-transport methods reduce this burden by reassigning fixed noise samples to data, but the source noise distribution itself remains passive. To address this, we introduce Contrastive Noise Alignment (CNA), a training-time method that creates dynamic, contrastive couplings by optimizing the noise representations directly. By modeling the noise batch as an interacting particle system, CNA employs a cross-modal InfoNCE objective to align noise particles with their paired data targets. To prevent spatial collapse, this alignment is regularized using an angular entropy term and a radial norm penalty. We show theoretically that this equilibrium asymptotically preserves Gaussian structures, maintaining tractability during inference. Empirically, CNA improves the alignment between noise and data, reduces flow curvature, and provides better generation quality with fewer required sampling steps. For few-step, pixel-space generation (2-4 NFEs), CNA reduces FID by over 50\% compared to standard rectified flow, and by at least 24\% against Optimal Transport baselines.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-17 in cs.LG).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.18488 — 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 2026-10-15T14:08:19.252Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-17?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).