|
--- |
|
license: apache-2.0 |
|
language: |
|
- en |
|
metrics: |
|
- roc_auc |
|
base_model: |
|
- EleutherAI/gpt-neo-2.7B |
|
pipeline_tag: text-classification |
|
tags: |
|
- machine-text-detection |
|
- machine-revised-text-detection |
|
--- |
|
|
|
<h1 align="center">Imitate Before Detect: Aligning Machine Stylistic Preference for Machine-Revised Text Detection</h1> |
|
|
|
<p align="center"> |
|
<a href="https://scholar.google.com/citations?user=Au_y5poAAAAJ">Jiaqi Chen</a><sup>*</sup>, <a href="https://xyzhu1225.github.io/">Xiaoye Zhu</a><sup>*</sup>, <a href="https://leolty.github.io/">Tianyang Liu</a><sup>*</sup>, Ying Chen, <a href="https://xinhuichen-02.github.io/">Xinhui Chen</a>,<br> <a href="https://scholar.google.com/citations?user=koA9QbMAAAAJ">Yiwen Yuan</a>, <a href="https://cooperleong00.github.io/">Chak Tou Leong</a>, <a href="https://zcli-charlie.github.io/">Zuchao Li</a><sup>†</sup>, Tang Long, <a href="https://yusalei.github.io/">Lei Zhang</a>, <br><a href="https://scholar.google.com/citations?user=281EWzQAAAAJ">Chenyu Yan</a>, <a href="https://scholar.google.com/citations?user=mliv6KEAAAAJ">Guanghao Mei</a>, <a href="https://scholar.google.com/citations?user=epTfECgAAAAJ">Jie Zhang</a><sup>†</sup>, <a href="https://scholar.google.com/citations?user=BLKHwNwAAAAJ">Lefei Zhang</a><sup>†</sup> |
|
</p> |
|
|
|
<p align="center"> |
|
*Equal contribution.<br> †Equal contribution of corresponding author. |
|
</p> |
|
|
|
Detecting **machine-revised text** remains a challenging task as it often involves subtle style changes embedded within human-originated content. The ImBD framework introduces a novel approach to tackle this problem, leveraging **style preference optimization (SPO)** and **Style-CPC** to effectively capture machine-style phrasing. Our method achieves state-of-the-art performance in detecting revisions by open-source and proprietary LLMs like GPT-3.5 and GPT-4o, demonstrating significant efficiency with minimal training data. |
|
|
|
We are excited to share our code and data to support further exploration in detecting machine-revised text. We welcome your feedback and invite collaborations to advance this field together! |
|
|
|
|
|
 |
|
|
|
## 🔥 News |
|
- **[2024, Dec 16]** Our online [demo](https://huggingface.co./spaces/machine-text-detection/ImBD) is available on hugging-face now! |
|
- **[2024, Dec 13]** Our [model](https://huggingface.co./xyzhu1225/ImBD/tree/main) and local inference code are available. |
|
- **[2024, Dec 9]** 🎉🎉 Our paper has been accepted by AAAI 25! |
|
- **[2024, Dec 7]** We've released our [website](https://machine-text-detection.github.io/ImBD). |
|
|
