SearchLLM: Detecting LLM Paraphrased Text by Measuring the Similarity with Regeneration of the Candidate Source via Search Engine
This addresses the challenge of detecting subtly paraphrased LLM-generated text for users and systems relying on content authenticity, though it appears incremental as it builds on existing detectors.
The paper tackles the problem of detecting LLM-paraphrased text that traditional methods often miss, by proposing SearchLLM, which uses search engines to find candidate sources and measures similarity with regenerated versions. Experimental results show it consistently enhances the accuracy of recent detectors in identifying such content and helps prevent paraphrasing attacks.
With the advent of large language models (LLMs), it has become common practice for users to draft text and utilize LLMs to enhance its quality through paraphrasing. However, this process can sometimes result in the loss or distortion of the original intended meaning. Due to the human-like quality of LLM-generated text, traditional detection methods often fail, particularly when text is paraphrased to closely mimic original content. In response to these challenges, we propose a novel approach named SearchLLM, designed to identify LLM-paraphrased text by leveraging search engine capabilities to locate potential original text sources. By analyzing similarities between the input and regenerated versions of candidate sources, SearchLLM effectively distinguishes LLM-paraphrased content. SearchLLM is designed as a proxy layer, allowing seamless integration with existing detectors to enhance their performance. Experimental results across various LLMs demonstrate that SearchLLM consistently enhances the accuracy of recent detectors in detecting LLM-paraphrased text that closely mimics original content. Furthermore, SearchLLM also helps the detectors prevent paraphrasing attacks.