Metraipolis › archive › dc_53826eaaf86e

Hierarchical Continuous Diffusion Language Models — arXiv:2610.02193

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.02193
License (as archived)
CC0-1.0
Archived
2026-10-06T12:16:31.437Z
SHA-256 of the archived text
d5a36251afd52155839d56d34268ca44fbb260e917ce99d085365a508168aa09
Stable URL of this page
https://metraipolis.com/documento/dc_53826eaaf86e

How to cite

Hierarchical Continuous Diffusion Language Models — arXiv:2610.02193 — arXiv (en), https://arxiv.org/abs/2610.02193 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_53826eaaf86e (Metraipolis archive dc_53826eaaf86e, sha256 d5a36251afd52155839d56d34268ca44fbb260e917ce99d085365a508168aa09)
BibTeX
@misc{dc53826eaaf86e,
  title = {Hierarchical Continuous Diffusion Language Models — arXiv:2610.02193},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.02193},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_53826eaaf86e; sha256 d5a36251afd52155839d56d34268ca44fbb260e917ce99d085365a508168aa09},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.02193. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Discrete diffusion language models offer a compelling alternative to autoregressive generation for tasks demanding bidirectional reasoning and global constraint satisfaction. Yet they share a structural bottleneck: when decoding in parallel, each token is sampled independently from its marginal, severing the statistical dependencies among the tokens decoded together. Continuous diffusion language models avoid this by denoising a shared continuous state, but their denoiser sees only that state, so nothing ties it to a valid token configuration until it is finally decoded. To address this, we propose Hierarchical Continuous Diffusion Language Models (HC-DLM), which couple discrete token generation with a continuous latent trajectory in a single, principled denoising process, whose training objective is derived from a variational bound on the token likelihood. In contrast to recent methods that attach continuous context to a self-contained discrete chain, HC-DLM makes the latent the only persistent generative state: tokens are read out from it at every step and feed back as a scaffold for the next latent update. On structured reasoning (Sudoku), mathematical planning (Countdown) and language modeling (LM1B), HC-DLM improves over discrete and continuous diffusion baselines at matched model size, in puzzle accuracy on Sudoku and Countdown and in generative perplexity on LM1B. Project page: this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-02 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.02193 — 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 2026-10-06T10:21:28.260Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).