Metraipolis › archive › dc_9f960f91f72a
Online Class Posteriors for Interpretable and Controllable Discrete Diffusion — arXiv:2611.03454
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.03454
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T11:23:17.994Z
- SHA-256 of the archived text
d4837d1857201fac5de29c3d2ee7b9899235c1863237ba7ce3dff759fd29cc29- Stable URL of this page
https://metraipolis.com/documento/dc_9f960f91f72a
How to cite
Online Class Posteriors for Interpretable and Controllable Discrete Diffusion — arXiv:2611.03454 — arXiv (en), https://arxiv.org/abs/2611.03454 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_9f960f91f72a (Metraipolis archive dc_9f960f91f72a, sha256 d4837d1857201fac5de29c3d2ee7b9899235c1863237ba7ce3dff759fd29cc29)
BibTeX
@misc{dc9f960f91f72a,
title = {Online Class Posteriors for Interpretable and Controllable Discrete Diffusion — arXiv:2611.03454},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.03454},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_9f960f91f72a; sha256 d4837d1857201fac5de29c3d2ee7b9899235c1863237ba7ce3dff759fd29cc29},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.03454. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Discrete diffusion models remain far less understood than their continuous counterparts, despite recent breakthroughs across modalities, most notably in text. Leveraging tools from continuous diffusion, we formulate a theoretical framework to track the class-conditional posterior in an online manner by combining conditional and unconditional model predictions. In masked language modeling, this posterior offers highly interpretable dynamics: tokens aligned with a class boost the posterior, while contradictory ones suppress it. We also show that the posterior can be used for logical guidance operations (AND, OR, NOT) in discrete diffusion. These operations enable precise control over generation, such as avoiding specific classes. In addition, we discuss a mismatch in the formulation of classifier-free guidance between discrete and continuous settings: discrete models mix the data distributions, while continuous models mix the marginals. We validate our methods and discuss their potential applications on synthetic data, with all quantities analytically tractable, as well as on MNIST, text, and antibody sequences. 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/2611.03454 — 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 2026-11-03T11:02:37.710Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).