Metraipolis › archive › dc_893d746dd6fc

Can Bayesian Optimization Efficiently Find a Strong Single Expert in Neural Thickets? — arXiv:2608.10867

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.10867
License (as archived)
CC0-1.0
Archived
2026-10-19T06:19:28.649Z
SHA-256 of the archived text
bac7b01561e323a115f4a21764371afa7b55caba299f6ae45238122e9c21c672
Stable URL of this page
https://metraipolis.com/documento/dc_893d746dd6fc

How to cite

Can Bayesian Optimization Efficiently Find a Strong Single Expert in Neural Thickets? — arXiv:2608.10867 — arXiv (en), https://arxiv.org/abs/2608.10867 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_893d746dd6fc (Metraipolis archive dc_893d746dd6fc, sha256 bac7b01561e323a115f4a21764371afa7b55caba299f6ae45238122e9c21c672)
BibTeX
@misc{dc893d746dd6fc,
  title = {Can Bayesian Optimization Efficiently Find a Strong Single Expert in Neural Thickets? — arXiv:2608.10867},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.10867},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_893d746dd6fc; sha256 bac7b01561e323a115f4a21764371afa7b55caba299f6ae45238122e9c21c672},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.10867. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Gradient-free post-training has emerged as a compelling alternative to gradient-based optimization for large language models (LLMs), but existing approaches remain costly. We ask whether structured search can identify a strong single expert under a modest evaluation budget. Motivated by evidence that useful weight updates lie in low-dimensional subspaces, we apply Bayesian optimization within a random linear embedding of weight space. Our method requires no backpropagation and uses a Gaussian process surrogate to guide candidate evaluations efficiently. Across several reasoning benchmarks with Qwen2.5-Instruct models from 0.5B to 3B parameters, Bayesian optimization using five times less candidate evaluations matches or exceeds RandOpt. These results show that surrogate-guided search can substantially reduce the evaluation cost of gradient-free post-training while producing stronger deployable single experts.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-12 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.10867 — 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 2026-10-19T06:11:45.097Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-12?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).