Metraipolis › archive › dc_7a7686f9beec

Bayesian Optimization with Rich Auxiliary Information via LLMs — arXiv:2609.19437

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.19437
License (as archived)
CC0-1.0
Archived
2026-10-10T04:08:38.018Z
SHA-256 of the archived text
dd8ab425bc4f48a39e9d1b16a5106351f875644241afdde74cea3d57ed0fc388
Stable URL of this page
https://metraipolis.com/documento/dc_7a7686f9beec

How to cite

Bayesian Optimization with Rich Auxiliary Information via LLMs — arXiv:2609.19437 — arXiv (en), https://arxiv.org/abs/2609.19437 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7a7686f9beec (Metraipolis archive dc_7a7686f9beec, sha256 dd8ab425bc4f48a39e9d1b16a5106351f875644241afdde74cea3d57ed0fc388)
BibTeX
@misc{dc7a7686f9beec,
  title = {Bayesian Optimization with Rich Auxiliary Information via LLMs — arXiv:2609.19437},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.19437},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7a7686f9beec; sha256 dd8ab425bc4f48a39e9d1b16a5106351f875644241afdde74cea3d57ed0fc388},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.19437. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Bayesian Optimization (BO) is widely used for optimizing expensive black-box functions, yet it typically reduces each expensive experiment to an input and its objective value, discarding much of the information the experiment produces. Such auxiliary information can include training curves in hyperparameter optimization, expert notes and images in scientific experimentation, or known facts about the specific optimization problem. Leveraging the ability of large language models (LLMs) to process diverse and unstructured information, we introduce two methods: AuxBO-Evolve, which uses auxiliary information to evolve beliefs over the location of the optimum, and AuxBO-Vicinity, which uses it to locally guide the acquisition function. Across synthetic tasks, hyperparameter optimization benchmarks, and a nuclear fusion task with unstructured scientist-written logs, our methods consistently improve optimization performance and outperform standard BO and existing LLM-based approaches. We further find that LLMs provide more effective probabilistic guidance when modeling likely maximizer locations rather than objective values pointwise. Overall, our results demonstrate that exploiting rich information beyond standard $(\mathbf{x}, y)$ interactions can substantially improve BO performance.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-18 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.19437 — 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 2026-10-10T03:48:18.258Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).