Zero-Shot Detection of LLM-Generated Text using Token Cohesiveness
S.L. Ma, Quan Wang · 2024
The increasing capability and widespread usage of large language models (LLMs) highlight the desirability of automatic detection of LLMgenerated text.Zero-shot detectors, due to their training-free nature, have received considerable attention and notable success.In this paper, we identify a new feature, token cohesiveness, that is useful for zero-shot detection, and we demonstrate that LLM-generated text tends to exhibit higher token cohesiveness than human-written text.Based on this observation, we devise TOC-SIN, a generic dual-channel detection paradigm that uses token cohesiveness as a plug-and-play module to improve existing zero-shot detectors.To calculate token cohesiveness, TOCSIN only requires a few rounds of random token deletion and semantic difference measurement, making it particularly suitable for a practical black-box setting where the source model used for generation is not accessible.Extensive experiments with four state-of-the-art base detectors on various datasets, source models, and evaluation settings demonstrate the effectiveness and generality of the proposed approach.Code available at: https://github.com/Shixuan-Ma/TOCSIN.