Metraipolis › archive › dc_5c906eb573cd
Amortizing Sequential Experimental Design in Reasoning LLMs — arXiv:2610.20832
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.20832
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T09:06:30.690Z
- SHA-256 of the archived text
07b85b82280ed6b93a13cb3696839bfe3915017e5826ddeed8681b872a3304dc- Stable URL of this page
https://metraipolis.com/documento/dc_5c906eb573cd
How to cite
Amortizing Sequential Experimental Design in Reasoning LLMs — arXiv:2610.20832 — arXiv (en), https://arxiv.org/abs/2610.20832 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5c906eb573cd (Metraipolis archive dc_5c906eb573cd, sha256 07b85b82280ed6b93a13cb3696839bfe3915017e5826ddeed8681b872a3304dc)
BibTeX
@misc{dc5c906eb573cd,
title = {Amortizing Sequential Experimental Design in Reasoning LLMs — arXiv:2610.20832},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.20832},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5c906eb573cd; sha256 07b85b82280ed6b93a13cb3696839bfe3915017e5826ddeed8681b872a3304dc},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.20832. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Despite their extensive pretraining knowledge and strong reasoning capabilities, large language models (LLMs) struggle to adaptively gather information in experimental design tasks. We study how to amortize sequential Bayesian experimental design (BED) into an LLM by explicitly training the reasoning that guides action selection using information-gain rewards. Our approach specifies the experimental generative model independently of the LLM's own beliefs, allowing us to optimize expected information gain under a model of our choosing. We investigate the design choices for amortization, including alternative generative-model constructions, reward estimation, hypothesis-space specification, multi-step objectives, and computational costs relative to training- and inference-time BED baselines. On Twenty Questions, the resulting policies improve information acquisition and task performance, outperform inference- and training-time BED baselines at substantially lower computational cost, and generalize to unseen categories. We further demonstrate positive transfer from Twenty Questions training to Medi-Q, a real-world clinical dialogue benchmark. These results suggest that information gathering can be learned as a generalizable reasoning skill in LLMs. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.20832 — 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 2026-10-16T08:44:17.874Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).