Saved in:
Bibliographic Details
Main Authors: Zhou, Yinghan, Wen, Juan, Peng, Wanli, Xue, Yiming, Zhang, Ziwei, Wu, Zhengxian
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2504.21019
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918104069570560
author Zhou, Yinghan
Wen, Juan
Peng, Wanli
Xue, Yiming
Zhang, Ziwei
Wu, Zhengxian
author_facet Zhou, Yinghan
Wen, Juan
Peng, Wanli
Xue, Yiming
Zhang, Ziwei
Wu, Zhengxian
contents The growing popularity of large language models has raised concerns regarding the potential to misuse AI-generated text (AIGT). It becomes increasingly critical to establish an excellent AIGT detection method with high generalization and robustness. However, existing methods either focus on model generalization or concentrate on robustness. The unified mechanism, to simultaneously address the challenges of generalization and robustness, is less explored. In this paper, we argue that robustness can be view as a specific form of domain shift, and empirically reveal an intrinsic mechanism for model generalization of AIGT detection task. Then, we proposed a novel AIGT detection method (DP-Net) via dynamic perturbations introduced by a reinforcement learning with elaborated reward and action. Experimentally, extensive results show that the proposed DP-Net significantly outperforms some state-of-the-art AIGT detection methods for generalization capacity in three cross-domain scenarios. Meanwhile, the DP-Net achieves best robustness under two text adversarial attacks. The code is publicly available at https://github.com/CAU-ISS-Lab/AIGT-Detection-Evade-Detection/tree/main/DP-Net.
format Preprint
id arxiv_https___arxiv_org_abs_2504_21019
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Kill two birds with one stone: generalized and robust AI-generated text detection via dynamic perturbations
Zhou, Yinghan
Wen, Juan
Peng, Wanli
Xue, Yiming
Zhang, Ziwei
Wu, Zhengxian
Computation and Language
Artificial Intelligence
The growing popularity of large language models has raised concerns regarding the potential to misuse AI-generated text (AIGT). It becomes increasingly critical to establish an excellent AIGT detection method with high generalization and robustness. However, existing methods either focus on model generalization or concentrate on robustness. The unified mechanism, to simultaneously address the challenges of generalization and robustness, is less explored. In this paper, we argue that robustness can be view as a specific form of domain shift, and empirically reveal an intrinsic mechanism for model generalization of AIGT detection task. Then, we proposed a novel AIGT detection method (DP-Net) via dynamic perturbations introduced by a reinforcement learning with elaborated reward and action. Experimentally, extensive results show that the proposed DP-Net significantly outperforms some state-of-the-art AIGT detection methods for generalization capacity in three cross-domain scenarios. Meanwhile, the DP-Net achieves best robustness under two text adversarial attacks. The code is publicly available at https://github.com/CAU-ISS-Lab/AIGT-Detection-Evade-Detection/tree/main/DP-Net.
title Kill two birds with one stone: generalized and robust AI-generated text detection via dynamic perturbations
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2504.21019