C-ReD: A Comprehensive Chinese Benchmark for AI-Generated Text Detection Derived from Real-World Prompts

[论文] C-ReD: A Comprehensive Chinese Benchmark for AI-Generated Text Detection Derived from Real-World Prompts ## 论文概要 **研究领域**: cs.CL, cs.AI **作者**: Chenxi Qing, Junxi Wu, Zheng Liu, Yixiang Qiu, Hongyao Yu, Bin Chen, Hao Wu, Shu-Tao Xia **发布时间**: 2026-04-13 **arXiv**: [2604.11796](https://arxiv.org/abs/2604.11796) ## 中文摘要 近期大语言模型能够生成高度流畅的文本内容。虽然它们为人类提供了便利，但也引入了各种风险，如网络钓鱼和学术不端。大量研究工作致力于开发检测AI生成文本的算法和构建相关数据集。然而，在中文语料库领域，仍存在模型多样性有限和数据同质化等挑战。本文提出C-ReD：一个基于真实提示的中文AI生成文本检测综合基准。实验表明C-ReD不仅支持可靠的域内检测，还支持对未见LLM和外部中文数据集的强泛化能力。 ## 原文摘要 Recently, large language models (LLMs) are capable of generating highly fluent textual content. While they offer significant convenience to humans, they also introduce various risks, like phishing and academic dishonesty. Numerous research efforts have been dedicated to developing algorithms for detecting AI-generated text and constructing relevant datasets. --- *自动采集于 2026-04-15* #论文 #arXiv #AI #小凯