Metraipolis › archive › dc_fe1890f2f0de

Rethinking Evaluation for Principled Generalization Using a Kripkean Challenge to Meaning as Use — arXiv:2610.35111

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.35111
License (as archived)
CC0-1.0
Archived
2026-10-27T08:12:12.867Z
SHA-256 of the archived text
46b34047c51d277e01416812f5908cf636479a2d14bb519dc340df1ffb392349
Stable URL of this page
https://metraipolis.com/documento/dc_fe1890f2f0de

How to cite

Rethinking Evaluation for Principled Generalization Using a Kripkean Challenge to Meaning as Use — arXiv:2610.35111 — arXiv (en), https://arxiv.org/abs/2610.35111 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fe1890f2f0de (Metraipolis archive dc_fe1890f2f0de, sha256 46b34047c51d277e01416812f5908cf636479a2d14bb519dc340df1ffb392349)
BibTeX
@misc{dcfe1890f2f0de,
  title = {Rethinking Evaluation for Principled Generalization Using a Kripkean Challenge to Meaning as Use — arXiv:2610.35111},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.35111},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fe1890f2f0de; sha256 46b34047c51d277e01416812f5908cf636479a2d14bb519dc340df1ffb392349},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.35111. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large language models operationalize an intuitive, view of meaning as use. We argue that it faces a rule-following challenge. Using Kripke's interpretation of Wittgenstein, we argue that finite histories of use do not determine a unique rule for future extension. This reframes the debate about OOD generalization but also about NLP evaluation; benchmarks that measure performance on fixed use cases or local robustness, we argue, do not by themselves establish semantic competence. To evaluate whether models preserve rules and concepts under principled domain extension, we propose to evaluate two capacities, central to this ability: abstraction and recursion; and we offer an experimental paradigm for testing both.

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/2610.35111 — 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 2026-10-27T07:56:00.252Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).