Metraipolis › archive › dc_f55aa953d4a2
Where Should Learning Live? Regret Analysis for Text, Parameter, and Joint Adaptation — arXiv:2611.09701
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.09701
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-09T22:42:01.274Z
- SHA-256 of the archived text
e1f1e9bef290b3ff9ee1d0d67d925870f356b1a0cc057e4ae57a214b7ba1f712- Stable URL of this page
https://metraipolis.com/documento/dc_f55aa953d4a2
How to cite
Where Should Learning Live? Regret Analysis for Text, Parameter, and Joint Adaptation — arXiv:2611.09701 — arXiv (en), https://arxiv.org/abs/2611.09701 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f55aa953d4a2 (Metraipolis archive dc_f55aa953d4a2, sha256 e1f1e9bef290b3ff9ee1d0d67d925870f356b1a0cc057e4ae57a214b7ba1f712)
BibTeX
@misc{dcf55aa953d4a2,
title = {Where Should Learning Live? Regret Analysis for Text, Parameter, and Joint Adaptation — arXiv:2611.09701},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.09701},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f55aa953d4a2; sha256 e1f1e9bef290b3ff9ee1d0d67d925870f356b1a0cc057e4ae57a214b7ba1f712},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.09701. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Language-model systems can select prompts and parameter adapters repeatedly during deployment, using feedback to guide later choices. Earlier choices may change the context or environment inherited by later configurations. We study selection among preconstructed candidates and compare cumulative loss with the best fixed prompt--parameter pair used from the beginning. This excess loss separates into three terms: selection regret within the available family, the gap from restricting that family, and the benefit or harm of the actual deployment history. Selection is evaluated using the history each candidate would have produced if used throughout, which separates candidate quality from the effects of switching. When loss functions are fixed independently of selector randomness, past choices affect loss for only a bounded number of rounds, and every candidate's own-history loss is revealed, limiting switches yields sublinear expected regret within the selected family. We also quantify the costs and additional assumptions of independent probing and deployment-only feedback. Controlled experiments show that fewer switches can either lower or raise deployed loss, depending on whether inherited histories are harmful or helpful. With Qwen3.5-9B on Spider, text-only selection has lower within-family regret than joint selection, yet produces more failures because its candidate family excludes stronger configurations. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (stat.ML) ; Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2611.09701 — 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 2026-11-09T22:33:47.179Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).