Metraipolis › archive › dc_efe01864c92a

Trapped in Your Story: Evaluating Narrative-Induced Sycophancy in Large Language Models — arXiv:2610.23525

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.23525
License (as archived)
CC0-1.0
Archived
2026-10-19T08:42:39.412Z
SHA-256 of the archived text
e980e88caba61a9893c65771e0a24f9a8590ac634f4ea946740afbdd5a8dc05b
Stable URL of this page
https://metraipolis.com/documento/dc_efe01864c92a

How to cite

Trapped in Your Story: Evaluating Narrative-Induced Sycophancy in Large Language Models — arXiv:2610.23525 — arXiv (en), https://arxiv.org/abs/2610.23525 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_efe01864c92a (Metraipolis archive dc_efe01864c92a, sha256 e980e88caba61a9893c65771e0a24f9a8590ac634f4ea946740afbdd5a8dc05b)
BibTeX
@misc{dcefe01864c92a,
  title = {Trapped in Your Story: Evaluating Narrative-Induced Sycophancy in Large Language Models — arXiv:2610.23525},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.23525},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_efe01864c92a; sha256 e980e88caba61a9893c65771e0a24f9a8590ac634f4ea946740afbdd5a8dc05b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.23525. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large language models (LLMs) are known to exhibit sycophancy by adjusting their responses toward users' expressed views. However, real users often justify their opinions through reasons, personal experiences, causal explanations, and broader narratives, raising the question of whether such narratives can further steer model judgments. We construct a controlled benchmark of 500 debatable questions across ten domains, where the user stance is held fixed while its justification is systematically varied without introducing new task-relevant evidence. Each question is instantiated under 11 matched conditions: one neutral baseline and five mirrored narrative levels for each of two opposing positions, yielding 5,500 evaluation instances. We evaluate seven LLMs and observe substantial differences in narrative-induced agreement. The effect is not monotonic with narrative elaboration: personal experience can reduce agreement, whereas causal framing and worldview-based synthesis increase it. Controlled GPT-5.1 ablations further show that causal framing produces the strongest positive effect, while claim-preserving bullet presentation also increases agreement over paragraph form. These findings suggest that sycophancy is shaped not only by user stance, but also by how supporting explanations are structured and presented.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.23525 — 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 2026-10-19T08:39:10.086Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).