用句子结构检测AI生成文本，抗改写、轻量高效。
Trace Is In Sentences: Unbiased Lightweight ChatGPT-Generated Text Detector
- 基于句子内部结构而非词语，避免被改写绕过
- 在两个数据集上准确率超90%，且对修改文本保持稳定
- 适合需要快速部署的反作弊系统或内容审核场景
ChatGPT的广泛使用引发对其滥用的担忧，亟需可靠的AI生成文本检测方法。现有词级检测器易受改写或简单提示（PSP）影响，受模型自身词级模式（CWP）和训练数据偏见干扰，对修改文本性能下降，且常需大型模型或在线LLM交互。为此，我们提出新任务：检测原始及经PSP修改的AI生成文本，并设计轻量级框架，通过编码预训练语言模型的句子嵌入并建模其关系实现分类，其内部结构在词级变化下保持不变。采用对比学习缓解自回归生成带来的嵌入偏见，并引入因果图与反事实方法，剥离主题相关偏见，分离出结构性特征。在两个精心构建的数据集（包括摘要对比和修订版生活问答）上的实验验证了该方法的有效性。
原文摘要 · Abstract (English)
The widespread adoption of ChatGPT has raised concerns about its misuse, highlighting the need for robust detection of AI-generated text. Current word-level detectors are vulnerable to paraphrasing or simple prompts (PSP), suffer from biases induced by ChatGPT's word-level patterns (CWP) and training data content, degrade on modified text, and often require large models or online LLM interaction. To tackle these issues, we introduce a novel task to detect both original and PSP-modified AI-generated texts, and propose a lightweight framework that classifies texts based on their internal structure, which remains invariant under word-level changes. Our approach encodes sentence embeddings from pre-trained language models and models their relationships via attention. We employ contrastive learning to mitigate embedding biases from autoregressive generation and incorporate a causal graph with counterfactual methods to isolate structural features from topic-related biases. Experiments on two curated datasets, including abstract comparisons and revised life FAQs, validate the effectiveness of our method.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。