Metraipolis › archive › dc_9dca06642d10

Self-Conditioning Enables Deep Sequential Reasoning in Flow Matching — arXiv:2610.19628

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.19628
License (as archived)
CC0-1.0
Archived
2026-10-15T14:23:31.283Z
SHA-256 of the archived text
65ba1e537a549c0eb7a46c0bc5fa6c29eab9b5fa59d6d58727d2f6b83d6ec008
Stable URL of this page
https://metraipolis.com/documento/dc_9dca06642d10

How to cite

Self-Conditioning Enables Deep Sequential Reasoning in Flow Matching — arXiv:2610.19628 — arXiv (en), https://arxiv.org/abs/2610.19628 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_9dca06642d10 (Metraipolis archive dc_9dca06642d10, sha256 65ba1e537a549c0eb7a46c0bc5fa6c29eab9b5fa59d6d58727d2f6b83d6ec008)
BibTeX
@misc{dc9dca06642d10,
  title = {Self-Conditioning Enables Deep Sequential Reasoning in Flow Matching — arXiv:2610.19628},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.19628},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_9dca06642d10; sha256 65ba1e537a549c0eb7a46c0bc5fa6c29eab9b5fa59d6d58727d2f6b83d6ec008},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.19628. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Recent work has explored visual reasoning through image generation, where models solve visual-reasoning problems by directly generating solution images. Intuitively, iterative samplers such as flow matching are well suited to this task, since each sampling step may resemble one reasoning step. However, in this work, from a circuit complexity perspective, we show that this intuition is misleading. We prove that as long as the model outputs satisfy certain mild regularity conditions, the trajectory of the vanilla flow-matching sampler can be simulated by a constant number of sequential rounds of parallel model evaluations. Consequently, additional sampling steps cannot increase the expressive power of a constant-depth backbone model under the standard complexity-theoretic model of transformers. We then reveal that self-conditioning, which conditions the model on the previous step's estimate of the final image, plays a previously unrecognized role in removing this bottleneck. Controlled experiments on diverse visual-reasoning tasks show that flow matching with self-conditioning achieves higher accuracy and greater reasoning depth than vanilla flow matching, and its performance improves with more sampling steps. Theoretically, we prove that with self-conditioning, a constant-depth backbone can achieve expressive power equivalent to that of space-bounded Turing machines. Our study suggests a design principle for future visual reasoning research: carry more information between sampling steps, so that additional steps translate into deeper reasoning.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.19628 — 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 2026-10-15T13:53:35.887Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).