Metraipolis › archive › dc_44910e7177db

Verifiable, Articulable, and Tacit Components of Preference — arXiv:2610.03025

Source
arXiv
Original
https://arxiv.org/abs/2610.03025
License (as archived)
CC0-1.0
Archived
2026-10-05T22:49:27.740Z
SHA-256 of the archived text
e1e39ce4f7c012e3dd692157c8649db220af58e455af886a639103fcd5444a3d
Stable URL of this page
https://metraipolis.com/documento/dc_44910e7177db

How to cite

Verifiable, Articulable, and Tacit Components of Preference — arXiv:2610.03025 — arXiv, https://arxiv.org/abs/2610.03025 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_44910e7177db (Metraipolis archive dc_44910e7177db, sha256 e1e39ce4f7c012e3dd692157c8649db220af58e455af886a639103fcd5444a3d)
BibTeX
@misc{dc44910e7177db,
  title = {Verifiable, Articulable, and Tacit Components of Preference — arXiv:2610.03025},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.03025},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_44910e7177db; sha256 e1e39ce4f7c012e3dd692157c8649db220af58e455af886a639103fcd5444a3d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.03025. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

What makes a short story gripping; a news article newsworthy; or a math proof elegant? These constructs resist articulation or verification; their meaning is at least partially tacit. However, modern AI models are improved primarily via articulated constitutions, rubrics and verifiers (i.e. in RLAIF and RLVR); tacit components of preferences are typically understudied. We introduce a large, labeled preference dataset CreativePreferences, containing 2.8M texts labeled by 317M human preference judgments across 7 creative domains, with 42 benchmark tasks. We model these labels with executable programs, rubric banks and densely trained models (V, A and VAT, respectively). We observe robust articulability gaps, VAT-VA; and verifiability gaps, VAT-V; we estimate upper and lower bounds for each gap with a novel measurement approach that discovers articulable and verifiable metrics, identifies spurious variables and estimates the value of undiscovered metrics using capture-recapture. These gaps occur across all domains, even in domains traditionally treated as fully verifiable: correctness-centered domains (i.e. mathematics and software engineering) and claim- and novelty-centric domains (i.e. news, patents, peer review). The size of the gap varies based on domain (e.g. peer review and creative writing have the largest articulability gaps) and widens as more people take part in the judgment, consistent with Collins' collective tacit knowledge. We show two consequences: (1) on human generations, the full model more closely matches human preferences, often in disagreement with articulated criteria, and (2) in an analogy to Goodhart's law, articulating preference shifts it away from the tacit dimension. Articulability and verifiability gaps are consequential; we give recommendations on when tasks can be prompted; how learning mechanisms might improve; and when to leave judgments with humans.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL); Computers and Society (cs.CY); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.03025 — 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 2026-10-05T22:45:55.667Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).