Metraipolis › archive › dc_7efec13aa1e5

Training Autoregressive Language Models on Rounded, Bounded, and Censored Numerical Labels via Exact Likelihood — arXiv:2611.02461

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.02461
License (as archived)
CC0-1.0
Archived
2026-11-03T18:37:24.400Z
SHA-256 of the archived text
681bfbff1c6223c5be75388b50497a4c48ca392658aa0a9803244f210c3be0a8
Stable URL of this page
https://metraipolis.com/documento/dc_7efec13aa1e5

How to cite

Training Autoregressive Language Models on Rounded, Bounded, and Censored Numerical Labels via Exact Likelihood — arXiv:2611.02461 — arXiv (en), https://arxiv.org/abs/2611.02461 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7efec13aa1e5 (Metraipolis archive dc_7efec13aa1e5, sha256 681bfbff1c6223c5be75388b50497a4c48ca392658aa0a9803244f210c3be0a8)
BibTeX
@misc{dc7efec13aa1e5,
  title = {Training Autoregressive Language Models on Rounded, Bounded, and Censored Numerical Labels via Exact Likelihood — arXiv:2611.02461},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.02461},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7efec13aa1e5; sha256 681bfbff1c6223c5be75388b50497a4c48ca392658aa0a9803244f210c3be0a8},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.02461. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Numerical labels often specify a set of possible values: a measurement is rounded, a laboratory test reports only a bound, or a program is stopped at a timeout. We introduce boundary marginalization, the first method that trains an autoregressive language model on the exact likelihood of such observations through its existing digit tokens. A fixed-width, order-preserving numeral representation groups consecutive grid values under shared digit prefixes. An interval is a union of such groups, so its exact probability and gradient follow from at most two boundary paths, the successive digit prefixes of its endpoints, each of length logarithmic in the number of grid values. The resulting objective unifies precise, rounded, bounded, and censored supervision and trains digit decisions at the resolution supplied by each observation. The same computation supports exact distributional queries and unbiased gradients for the continuous ranked probability score on exact labels. Experiments show recovery of a conditional distribution from rounded and censored reports alone, improved likelihood on naturally censored molecular measurements, and more accurate code latency distributions on two backbones. Exact interval updates remain tractable as the numerical support grows to one million values.

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/2611.02461 — 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 2026-11-03T18:12:54.573Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).