TY  - RPRT
TI  - Learning from Response not Preference: A Stackelberg Approach for LLM Detoxification using Non-parallel Data
AU  - Xinhong Xie
AU  - Tao Li
AU  - Quanyan Zhu
PY  - 2024
UR  - https://arxiv.org/abs/2410.20298
ID  - 2410.20298
ER  -