Metraipolis › archive › dc_3e5fa4c6d796
Who Wrote It Is Not Enough: Detecting Who Contributed the Insight — arXiv:2610.07365
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.07365
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T11:23:03.013Z
- SHA-256 of the archived text
3cc74347e69ded7db358aef08d92f82ad9d0a2acb7e57ad4ba28f0374e1457b8- Stable URL of this page
https://metraipolis.com/documento/dc_3e5fa4c6d796
How to cite
Who Wrote It Is Not Enough: Detecting Who Contributed the Insight — arXiv:2610.07365 — arXiv (en), https://arxiv.org/abs/2610.07365 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3e5fa4c6d796 (Metraipolis archive dc_3e5fa4c6d796, sha256 3cc74347e69ded7db358aef08d92f82ad9d0a2acb7e57ad4ba28f0374e1457b8)
BibTeX
@misc{dc3e5fa4c6d796,
title = {Who Wrote It Is Not Enough: Detecting Who Contributed the Insight — arXiv:2610.07365},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.07365},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3e5fa4c6d796; sha256 3cc74347e69ded7db358aef08d92f82ad9d0a2acb7e57ad4ba28f0374e1457b8},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.07365. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
As LLMs increasingly assist scientific writing and peer review, detecting who wrote the text is no longer sufficient: we need to determine who contributed the underlying insight. We introduce Insight Provenance, the task of identifying whether a review insight originates from a human, an LLM, or their hybrid contribution. We construct InsightProv-v0 from 4,057 scientific papers and 12,660 human reviews, simulating different levels of LLM involvement with GPT-4o, Gemini, and DeepSeek and annotating provenance at the sentence level. We show that strong performance on raw data can be misleading, as models exploit linguistic and textual-authorship shortcuts that degrade substantially under progressively debiased evaluation. We therefore propose a two-stage adversarial framework that suppresses shortcut signals while preserving provenance-relevant information. Beyond detection, extensive analyses reveal what makes intellectual authorship identifiable: paper grounding and neighboring review context provide complementary provenance signals, while human, hybrid, and AI insights systematically differ in their information sources and failure modes. Most strikingly, AI insights predominantly remain close to generic or paper-provided information, whereas human insights more often introduce external knowledge and independent judgment. These findings suggest that while wording can be rewritten by an LLM, the provenance of an idea leaves a deeper and more persistent signal. 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/2610.07365 — 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 2026-10-07T11:15:39.415Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).