Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wei, Dongjun, Mao, Minjia, Fang, Xiao, Chau, Michael
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2504.02873
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912307947241472
author Wei, Dongjun
Mao, Minjia
Fang, Xiao
Chau, Michael
author_facet Wei, Dongjun
Mao, Minjia
Fang, Xiao
Chau, Michael
contents The malicious usage of large language models (LLMs) has motivated the detection of LLM-generated texts. Previous work in topological data analysis shows that the persistent homology dimension (PHD) of text embeddings can serve as a more robust and promising score than other zero-shot methods. However, effectively detecting short LLM-generated texts remains a challenge. This paper presents Short-PHD, a zero-shot LLM-generated text detection method tailored for short texts. Short-PHD stabilizes the estimation of the previous PHD method for short texts by inserting off-topic content before the given input text and identifies LLM-generated text based on an established detection threshold. Experimental results on both public and generated datasets demonstrate that Short-PHD outperforms existing zero-shot methods in short LLM-generated text detection. Implementation codes are available online.
format Preprint
id arxiv_https___arxiv_org_abs_2504_02873
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Short-PHD: Detecting Short LLM-generated Text with Topological Data Analysis After Off-topic Content Insertion
Wei, Dongjun
Mao, Minjia
Fang, Xiao
Chau, Michael
Computation and Language
The malicious usage of large language models (LLMs) has motivated the detection of LLM-generated texts. Previous work in topological data analysis shows that the persistent homology dimension (PHD) of text embeddings can serve as a more robust and promising score than other zero-shot methods. However, effectively detecting short LLM-generated texts remains a challenge. This paper presents Short-PHD, a zero-shot LLM-generated text detection method tailored for short texts. Short-PHD stabilizes the estimation of the previous PHD method for short texts by inserting off-topic content before the given input text and identifies LLM-generated text based on an established detection threshold. Experimental results on both public and generated datasets demonstrate that Short-PHD outperforms existing zero-shot methods in short LLM-generated text detection. Implementation codes are available online.
title Short-PHD: Detecting Short LLM-generated Text with Topological Data Analysis After Off-topic Content Insertion
topic Computation and Language
url https://arxiv.org/abs/2504.02873