Decoupling Content and Expression: Two-Dimensional Detection of AI-Generated Text
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866929736337326080 |
|---|---|
| author | Bao, Guangsheng Rong, Lihua Zhao, Yanbin Zhou, Qiji Zhang, Yue |
| author_facet | Bao, Guangsheng Rong, Lihua Zhao, Yanbin Zhou, Qiji Zhang, Yue |
| contents | The wide usage of LLMs raises critical requirements on detecting AI participation in texts. Existing studies investigate these detections in scattered contexts, leaving a systematic and unified approach unexplored. In this paper, we present HART, a hierarchical framework of AI risk levels, each corresponding to a detection task. To address these tasks, we propose a novel 2D Detection Method, decoupling a text into content and language expression. Our findings show that content is resistant to surface-level changes, which can serve as a key feature for detection. Experiments demonstrate that 2D method significantly outperforms existing detectors, achieving an AUROC improvement from 0.705 to 0.849 for level-2 detection and from 0.807 to 0.886 for RAID. We release our data and code at https://github.com/baoguangsheng/truth-mirror. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_00258 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Decoupling Content and Expression: Two-Dimensional Detection of AI-Generated Text Bao, Guangsheng Rong, Lihua Zhao, Yanbin Zhou, Qiji Zhang, Yue Computation and Language Artificial Intelligence Computers and Society The wide usage of LLMs raises critical requirements on detecting AI participation in texts. Existing studies investigate these detections in scattered contexts, leaving a systematic and unified approach unexplored. In this paper, we present HART, a hierarchical framework of AI risk levels, each corresponding to a detection task. To address these tasks, we propose a novel 2D Detection Method, decoupling a text into content and language expression. Our findings show that content is resistant to surface-level changes, which can serve as a key feature for detection. Experiments demonstrate that 2D method significantly outperforms existing detectors, achieving an AUROC improvement from 0.705 to 0.849 for level-2 detection and from 0.807 to 0.886 for RAID. We release our data and code at https://github.com/baoguangsheng/truth-mirror. |
| title | Decoupling Content and Expression: Two-Dimensional Detection of AI-Generated Text |
| topic | Computation and Language Artificial Intelligence Computers and Society |
| url | https://arxiv.org/abs/2503.00258 |