Exploring Generalized Features For LLM-Generated Text Detection

doi:10.1007/978-981-95-3729-7_25

2026-01-01

PRE

DOI：10.1007/978-981-95-3729-7_25

原文链接

原文求助

摘要

En 中文

大型语言模型（LLMs）的快速进步使得区分LLM生成文本和人类撰写文本日益困难，引发了关于真实性和安全性的担忧。尽管监督训练可以为特定LLM生成有效的文本检测器，但新模型的不断涌现使得为每个LLM和应用场景标注数据并训练单独模型的过程不切实际。为解决此挑战，我们提出了一种利用通用特征的全新框架。具体而言，我们引入了LLM条件特征对齐（LCFA）来引导模型学习LLM生成文本的领域不变特征。此外，我们整合了动态对比学习（DCL）以增强模型对数据扰动的鲁棒性，从而提升学习表征的泛化能力。为便于在真实条件下进行评估，我们构建了一个新数据集MLS，包含多场景和多语言下最先进LLM生成的文本。在MLS数据集上的实验结果证明了我们提出方法的有效性。

Keyword:

LLM-Generated Text Detection

Domain Generalization

期刊

IMAGE AND GRAPHICS, ICIG 2025, PT III

IF:

论文数:

被引数:

机构

chinese academy of sciences

学者数:

56.7W

论文数: 45.0W

被引数: 704

引用论文

Don’t Give Me the Details, Just the Summary! Topic-Aware Convolutional Neural Networks for Extreme Summarization 不要给我细节，只是总结!面向极端摘要的主题感知卷积神经网络

Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing

IF0

2018-01-01

OAAI

Shashi Narayan; Shay B. Cohen; Mirella Lapata

Automatic Detection of Generated Text is Easiest when Humans are Fooled

Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics

IF0

2020-01-01

OAAI

Daphne Ippolito; Daniel Duckworth; Chris Callison-Burch; Douglas Eck

IF0

PREAI

On the Zero-Shot Generalization of Machine-Generated Text Detectors

Findings of the Association for Computational Linguistics: EMNLP 2023

IF0

2023-01-01

PREAI

Pu,Xiao; Zhang,Jingyu; Han,Xiaochuang; Tsvetkov,Yulia; He,Tianxing

MedDialog: Large-scale Medical Dialogue Datasets MedDialog: 大规模医学对话数据集

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

IF0

2020-01-01

OAAI

Guangtao Zeng; Wenmian Yang; Zeqian Ju; Yue Yang; Sicheng Wang; Ruisi Zhang; Meng Zhou; Jiaqi Zeng; Xiangyu Dong; Ruoyu Zhang; Hongchao Fang; Penghui Zhu; Shu Chen; Pengtao Xie

author=Wang, Yuxia, and Mansurov, Jonibek, and Ivanov, Petar, and Su, Jinyan, and Shelmanov, Artem, and Tsvigun, Akim, and Whitehouse, Chenxi, and Mohammeda Afzal, Osama, and Mahmoud, Tarek, and Sasaki, Toru, and Arnold, Thomas, and Aji, Alhama Fikri, and Habash, Nizar, and Gurevych, Iryna, and Nakov, Preslav,, editor=Graham, Yvette, and Purver, Matthew,, publisher=Association for Computational Linguistics , booktitle=Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), pages=1369–1407, title=M4: Multi-Generator, Multi-Domain, and Multi-Lingual Black-Box Machine-Generated Text Detection, location=St. Julian’s, Malta , year=2024-03,作者=Wang, Yuxia, and Mansurov, Jonibek, and Ivanov, Petar, and Su, Jinyan, and Shelmanov, Artem, and Tsvigun, Akim, and Whitehouse, Chenxi, and Mohammeda Afzal, Osama, and Mahmoud, Tarek, and Sasaki, Toru, and Arnold, Thomas, and Aji, Alhama Fikri, and Habash, Nizar, and Gurevych, Iryna, and Nakov, Preslav,编辑=Graham, Yvette, and Purver, Matthew,出版社=Association for Computational Linguistics,书名=Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers),页码=1369–1407,标题=M4: Multi-Generator, Multi-Domain, and Multi-Lingual Black-Box Machine-Generated Text Detection,地点=St. Julian’s, Malta,年份=2024-03

IF0

PREAI

MAGE: Machine-generated Text Detection in the Wild MAGE：开放场景下的机器生成文本检测

Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

IF0

2024-01-01

PREAI

Yafu Li; Qintong Li; Leyang Cui; Wei Bi; Zhilin Wang; Longyue Wang; Linyi Yang; Shuming Shi; Yue Zhang

ConDA: Contrastive Domain Adaptation for AI-generated Text Detection ConDA：用于AI生成文本检测的对比域自适应

Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)

IF0