Metraipolis › archive › dc_b91c4c81afbf

Training Language Models To Be Coherent Decision-Makers — arXiv:2610.09164

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.09164
License (as archived): CC0-1.0
Archived: 2026-10-08T11:09:03.177Z
SHA-256 of the archived text: 9fe9400d00bbdb75a828d41fd1a579f2d800e4071981d8ddd970290c48b53d87
Stable URL of this page: https://metraipolis.com/documento/dc_b91c4c81afbf

How to cite

Training Language Models To Be Coherent Decision-Makers — arXiv:2610.09164 — arXiv (en), https://arxiv.org/abs/2610.09164 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b91c4c81afbf (Metraipolis archive dc_b91c4c81afbf, sha256 9fe9400d00bbdb75a828d41fd1a579f2d800e4071981d8ddd970290c48b53d87)

BibTeX

@misc{dcb91c4c81afbf,
  title = {Training Language Models To Be Coherent Decision-Makers — arXiv:2610.09164},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.09164},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b91c4c81afbf; sha256 9fe9400d00bbdb75a828d41fd1a579f2d800e4071981d8ddd970290c48b53d87},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.09164. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Reliable decision-making requires more than accurate prediction: a model must preserve its beliefs, apply the relevant utilities, and recognize when the information needed to justify an action is missing. We study whether language models can learn this decision procedure from supervised fine-tuning and generalize it across domains and differing natural-language expressions of the decision challenge. Across 20 datasets, we explore challenges of belief instability and decision-making errors by first eliciting probabilities of outcomes and then varying only the utilities and the framing of the decision problems, while holding the evidence fixed. We train models to preserve elicited beliefs while selecting the action that maximizes expected utility, and evaluate transfer to unseen application domains, held-out framings, and different classes of payoff structures. We further introduce incomplete-information settings in which required utilities are withheld and replaced with irrelevant text, testing whether models can distinguish missing decision-relevant information from merely additional context. We find that targeted fine-tuning substantially improves coherent decision-making and that in many situations, learning transfers across domains and framings to situations unobserved during training. Further, models trained for decidability learn to identify when action cannot be justified based on missing information. Finally, we show the value of a routed system that considers separately the recognition of decision completeness and utility-sensitive decision execution.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.09164 — 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 2026-10-08T10:14:05.912Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).