Metraipolis › archive › dc_a313e7faf71e
Strangers to Themselves: What Language Models Say About Themselves Is Generic — arXiv:2609.09899
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.09899
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T07:26:36.280Z
- SHA-256 of the archived text
a4ea87e0a57745071733c527b4abdb0d575798c192c6f8c6917d9db61b49986a- Stable URL of this page
https://metraipolis.com/documento/dc_a313e7faf71e
How to cite
Strangers to Themselves: What Language Models Say About Themselves Is Generic — arXiv:2609.09899 — arXiv (en), https://arxiv.org/abs/2609.09899 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a313e7faf71e (Metraipolis archive dc_a313e7faf71e, sha256 a4ea87e0a57745071733c527b4abdb0d575798c192c6f8c6917d9db61b49986a)
BibTeX
@misc{dca313e7faf71e,
title = {Strangers to Themselves: What Language Models Say About Themselves Is Generic — arXiv:2609.09899},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.09899},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a313e7faf71e; sha256 a4ea87e0a57745071733c527b4abdb0d575798c192c6f8c6917d9db61b49986a},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.09899. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Language models can fluently describe how they would behave: whether they would cave to pushback, misuse a tool, or lie under pressure. Is that description actually about the model speaking? We turn self-knowledge into a prediction test. Across nine behavioral evaluations, we measure how a model behaves under different conditions, ask it to predict those rates, and compare its predictions with controls that remove the self from the question. We find that: (i) Direct self-report is weak (r = +0.04), and even showing the model the exact items only raises prediction to +0.24. Crucially, the same item-informed question about "capable AI agents in general" does just as well (+0.28), while other models' answers about themselves predict the target model at least as well as its own. (ii) Frontier scale does not detectably change this pattern: any gains in prediction are not self-specific, and are consistent with a better theory of how AI assistants behave rather than better self-knowledge. (iii) First-person framing does have one robust effect: it shifts reports in the flattering direction, understating harmful behavior relative to the same question about a generic agent. (iv) Finetuning on a model's own behavioral record can teach narrow self-predictions, but it also changes the behavior being predicted and the gains do not transfer broadly. The practical implication is simple: asking a model what it would do mostly reveals a theory of AI assistants in general, plus a favorable bias, rather than privileged knowledge of that model. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-10 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV); Computers and Society (cs.CY). Pagina del lavoro: https://arxiv.org/abs/2609.09899 — 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 2026-10-16T07:00:16.363Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-10?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).