东京都立大学论文：Pangram 漏检 79.8% 由 Meta Muse-Glimmer 改写的科学摘要
Pangram, the AI-text detector, missed 79.8% of scientific abstracts rewritten by Meta's Muse-Glimmer, while flagging just 1 of 5,000 human abstracts.
东京都立大学一篇新论文发现，AI 文本检测器 Pangram 漏掉了 79.8% 由 Meta 的 Muse-Glimmer 改写的科学摘要，同时只把 5000 篇人类摘要中的 1 篇误判为 AI 生成。Pangram 的漏检率主要取决于改写所用的 LLM 版本，它能识别 93.5% 的 GPT-5 改写内容。
正文
当前显示原文，中文全文译文尚未就绪。
Pangram, the AI-text detector, missed 79.8% of scientific abstracts rewritten by Meta's Muse-Glimmer, while flagging just 1 of 5,000 human abstracts.
In a new paper from Tokyo Metropolitan University, reseaerchers find the share of AI-rewritten abstracts that Pangram misses depends strongly on the LLM version
Shows that it caught 93.5% of GPT-5 rewrites but missed 79.8% from another new model. Its miss rate depended mostly on which model did the rewriting.