Metraipolis › archive › dc_94ee030adb05
How Language Models Learn Taste: Brand Preference in LLMs — arXiv:2610.28272
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.28272
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-21T07:09:06.506Z
- SHA-256 of the archived text
69e771e6b9c19fdf222e41bb011644adc39cd4025a279b158769923b7bb64938- Stable URL of this page
https://metraipolis.com/documento/dc_94ee030adb05
How to cite
How Language Models Learn Taste: Brand Preference in LLMs — arXiv:2610.28272 — arXiv (en), https://arxiv.org/abs/2610.28272 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_94ee030adb05 (Metraipolis archive dc_94ee030adb05, sha256 69e771e6b9c19fdf222e41bb011644adc39cd4025a279b158769923b7bb64938)
BibTeX
@misc{dc94ee030adb05,
title = {How Language Models Learn Taste: Brand Preference in LLMs — arXiv:2610.28272},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.28272},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_94ee030adb05; sha256 69e771e6b9c19fdf222e41bb011644adc39cd4025a279b158769923b7bb64938},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.28272. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Asked to name a running shoe brand, language models say Nike. Asked to recommend one, they say Brooks. This paper asks where models learn that taste. Across 102 models and 42 brand categories, the brand a model picks when recommending differs from the brand it names in 26. The field moves together to a shared recommendation. The Named brands are market leaders, and the Picked brands are mostly niche and specialist. Three open training pipelines show where this forms. Pretraining already attaches the recommended brand to advice. Supervised fine-tuning teaches the model to answer and mostly leaves the pick unchanged. Preference optimization (DPO) teaches the model which answer is better, and the pick moves in all three pipelines, by 11 to 24 points. The preference data contains almost no brands. DPO moves the model's judgments toward a second set of associations even when the preferred brand is not in the preference data, and the effect reaches beyond brands. Asked to choose a sauce, models switch from ketchup to pesto. We do not know exactly how these preferences arise. We suspect they come from the judge models that decide which answers count as better during preference training. For brands, a recommendation depends less on what a model knows about them than on whether they fit this taste. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2610.28272 — 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 2026-10-21T07:05:58.145Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).