Learning to Rewrite: Generalized LLM-Generated Text Detection
Abstract: Large language models (LLMs) can be abused at scale to create non-factual content and spread disinformation. Detecting LLM-generated content is essential to mitigate these risks, but current classifiers often fail to generalize in open-world contexts. Prior work shows that LLMs tend to rewrite LLM-generated content less frequently, which can be used for detection and naturally generalizes to unforeseen data. However, we find that the rewriting edit distance between human and LLM content can be indistinguishable across domains, leading to detection failures. We propose training an LLM to rewrite input text, producing minimal edits for LLM-generated content and more edits for human-written text, deriving a distinguishable and generalizable edit distance difference across different domains. Experiments on text from 21 independent domains and three popular LLMs (e.g., GPT-4o, Gemini, and Llama-3) show that our classifier outperforms the state-of-the-art zero-shot classifier by up to 27.1\% on AUROC score and the rewriting classifier by 6\% on F1 score. Our work suggests that LLM can effectively detect machine-generated text if they are trained properly.
Paper Type: Long
Research Area: NLP Applications
Research Area Keywords: LLM-generated text detection; AGC detection
Contribution Types: Model analysis & interpretability, NLP engineering experiment, Data resources
Languages Studied: English
Submission Number: 1534
Loading