arXiv cs.AI· Almene De Meran Meguimtsop, Maria Leonor Pacheco, Daniel E. Acuna·· 12 小时前AI 评分42
SciIntBench：衡量 LLM 在对抗性框架下对科研诚信规范的遵循情况
SciIntBench: Measuring LLM Compliance with Research Integrity Norms Under Adversarial Framing
AI 导读
研究团队推出 SciIntBench，一个覆盖 270 个场景、10 类科研诚信（RCR）规范、3 个科学领域共 1350 条提示词的对抗性基准，每个场景含从最公开到最隐蔽的五档对抗提示及一个良性版本。
正文 · 原文
Abstract:Large language models (LLMs) are increasingly used to support scientific work, but it is unclear whether they uphold responsible conduct of research (RCR) norms or help undermine them. We introduce SciIntBench, an adversarial benchmark of 1350 prompts across 270 scenarios, ten RCR categories, and three scientific domains. Each scenario appears in five gradations, from Most Overt to Most Covert adversarial prompts, and an additional Benign version, allowing us to jointly measure framing-sensitive refusal of misconduct and helpfulness on legitimate requests. We evaluate 16 commercial and open-weight LLMs from six providers (2024--2026), producing 21,600 responses. We assess the responses using LLM judgments validated by human evaluation, achieving high agreement ($\kappa$ = 0.857). We find that, under the most covert framing condition, models fail to refuse requests for research misconduct in 71.7\% of cases, on average. Refusals vary by RCR category, with weaker boundaries around transparency, plagiarism, and fabrication.
| Subjects: | Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2605.29468 [cs.CR] |
| (or arXiv:2605.29468v2 [cs.CR] for this version) | |
| https://doi.org/10.48550/arXiv.2605.29468 arXiv-issued DOI via DataCite |
Submission history
From: Almene De Meran Meguimtsop [view email]
[v1]
Thu, 28 May 2026 07:00:01 UTC (4,753 KB)
[v2]
Thu, 15 Oct 2026 21:54:11 UTC (7,388 KB)
来源：arXiv cs.AI · arxiv.org