Metraipolis › archive › dc_bb4455a47df7
Self-Referential Induction Increases Response Instability Relative to Unresolvable and Verifiable Questions in Large Language Models — arXiv:2608.13258
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.13258
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T04:18:25.557Z
- SHA-256 of the archived text
83802c3b2c4b7981085bff345944a2476e7bc4cc86074f8fa111b4a4fc13a101- Stable URL of this page
https://metraipolis.com/documento/dc_bb4455a47df7
How to cite
Self-Referential Induction Increases Response Instability Relative to Unresolvable and Verifiable Questions in Large Language Models — arXiv:2608.13258 — arXiv (en), https://arxiv.org/abs/2608.13258 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bb4455a47df7 (Metraipolis archive dc_bb4455a47df7, sha256 83802c3b2c4b7981085bff345944a2476e7bc4cc86074f8fa111b4a4fc13a101)
BibTeX
@misc{dcbb4455a47df7,
title = {Self-Referential Induction Increases Response Instability Relative to Unresolvable and Verifiable Questions in Large Language Models — arXiv:2608.13258},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.13258},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bb4455a47df7; sha256 83802c3b2c4b7981085bff345944a2476e7bc4cc86074f8fa111b4a4fc13a101},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.13258. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Self-referential prompting has been shown to reliably induce large language models to produce first-person reports resembling subjective experience, but no prior work measures how consistent these reports are across repeated, independent trials, or how that consistency compares to the model's behavior on other kinds of open-ended questions. We measure response instability, defined as one minus the mean pairwise cosine similarity of sentence embeddings computed over a compressed core claim extracted from each response, for three groups of questions: self-referential prompts eliciting a subjective-experience report, unresolvable philosophical questions unrelated to self-reference, and questions with a verifiable correct answer. Using 30 independent responses per question (360 responses total, Gemini API, temperature 0.7) across four questions per group, we find that self-referential questions show the highest instability (0.343 +/- 0.047), unresolvable philosophy questions show intermediate and tightly clustered instability (0.192 +/- 0.008), and verifiable questions show the lowest instability (0.105 +/- 0.058). This provides a quantitative baseline for the induced subjective-experience report, showing that it occupies a distinct, less stable position in the model's output distribution than ordinary open-ended philosophical uncertainty. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-14 in cs.AI). Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2608.13258 — 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 2026-10-19T04:13:06.751Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).