arXiv：cs.CL· Zhuoyang Zou, Abolfazl Ansari, Jiaxi Yang, Delvin Ce Zhang, Qian Chen, Dongwon Lee, Wenpeng Yin·· 6 小时前AI 评分43
不止是谁写的：Insight Provenance 任务识别评审洞见来自人类、LLM 还是混合贡献
Who Wrote It Is Not Enough: Detecting Who Contributed the Insight
AI 导读
研究者提出 Insight Provenance 任务，判断评审洞见源自人类、LLM 还是二者混合，并基于 4,057 篇科学论文和 12,660 条人类评审构建了 InsightProv-v0 数据集，用 GPT-4o、Gemini 和 DeepSeek 模拟不同 LLM 参与程度并进行句子级标注。
正文
Abstract: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.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.07365 [cs.CL] |
| (or arXiv:2610.07365v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.07365 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zhuoyang Zou [view email]
[v1]
Mon, 5 Oct 2026 20:34:03 UTC (94 KB)
来源：arXiv：cs.CL · arxiv.org