IBM/RADAR
Code for our NeurIPS2023 accepted paper: RADAR: Robust AI-Text Detection via Adversarial Learning. We tested RADAR on 8 LLMs including Vicuna and LLaMA. The results show that RADAR can attain good detection performance on LLM-generated AI-text while being robust against paraphrasing.
Jupyter NotebookApache-2.0
Stargazers
- 202304703728
- 5oclockshadow
- AlbertRohUSC Marshall School of Business
- articlespinnerai
- Blue0rigin
- devanshkhandekarIndia
- electronicjude
- enhaohuang@Zhejiang University
- farawayxxx
- FloofCatResearcher
- GregxmHuThe Chinese University of Hong Kong
- gulldan
- klyedKLYE INC.
- kpedrok@mightyhealth
- kspicer80
- Loganll23
- Lyu6PosHaoPeking University
- NingNing0111NUN
- oktaydbk54Bekchilab
- oscaromsnPindograma
- PinkExHarbin Institute of Technology, Weihai
- refex
- s4nt0sHouston, Texas
- SajjjadddMy Room
- shyammarjitVCL @IISC
- SjorsLockhorstEnjins
- smoothdvd
- songhao664
- tmohn@aethr-xyz
- tom-doerr
- wangjing0cambridge, MA
- wuweh
- wwbs4564Renmin University of China
- yoke1990
- zhangwengyu999The Hong Kong Polytechnic University, Department of Computing
- zhuaizhuaiyeer