Metraipolis › archive › dc_f9bb7737eef7
Personalized Alignment of Language Models with Human Value Profiles — arXiv:2611.09919
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.09919
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-10T18:41:59.551Z
- SHA-256 of the archived text
70b9905908093b8b415f59f09be54b44e99d7dbb1d1a4ac7b569a4647c53c8e6- Stable URL of this page
https://metraipolis.com/documento/dc_f9bb7737eef7
How to cite
Personalized Alignment of Language Models with Human Value Profiles — arXiv:2611.09919 — arXiv (en), https://arxiv.org/abs/2611.09919 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f9bb7737eef7 (Metraipolis archive dc_f9bb7737eef7, sha256 70b9905908093b8b415f59f09be54b44e99d7dbb1d1a4ac7b569a4647c53c8e6)
BibTeX
@misc{dcf9bb7737eef7,
title = {Personalized Alignment of Language Models with Human Value Profiles — arXiv:2611.09919},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.09919},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f9bb7737eef7; sha256 70b9905908093b8b415f59f09be54b44e99d7dbb1d1a4ac7b569a4647c53c8e6},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.09919. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large language models are increasingly used for subjective tasks such as seeking advice, where there is often no single correct answer. Yet dominant alignment methods learn a single behavior policy that applies uniformly across users, limiting their ability to reflect individual value priorities. We propose personalized value alignment, a framework for adapting model responses to users' stable value profiles rather than to personas, demographics, or task-specific preferences used in prior personalization work. Our approach represents users through Schwartz value profiles and selects which of their values are relevant to each query before applying targeted prompting, activation steering, or LoRA-DPO. We create a dataset of 1,501 value-tension scenarios and evaluate over 152 diverse value profiles sampled from a nationally representative survey. Value-targeted methods outperform no-personalization, demographic, and full-profile prompting baselines, indicating that gains come from focused conditioning on query-specific value priorities rather than mere exposure. We further find that steering effects are not uniform across users: alignment gains vary by where a user's priorities fall in the value space. Combining targeted prompting with activation steering or LoRA-DPO yields the strongest personalization while keeping logical quality within adequate range. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Human-Computer Interaction (cs.HC) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2611.09919 — 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 2026-11-10T18:37:37.558Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).