Document is current

Article History

Received: 23 October 2025

Accepted: 28 May 2026

First Online: 18 June 2026

Declarations

:

: The primary outcomes derive from programmatic interaction with a publicly available LLM (Mistral-7B-Instruct-v0.3). Human involvement was limited to paid annotation of model outputs under written informed consent. An internal ethics review was conducted through the Ryan Research Institute; under applicable French and institutional guidelines, formal ethics approval was not required. Three human annotators participated as raters. Each provided written informed consent prior to annotation, following the protocol documented in docs/rater_protocol.md of the reproducibility package.

: All raters consented to the publication of anonymized annotation data. No personally identifiable information is included in the dataset or manuscript.

: The author declares no competing financial or non-financial interests.