TY  - RPRT
TI  - BEEAR: Embedding-based Adversarial Removal of Safety Backdoors in Instruction-tuned Language Models
AU  - Yi Zeng
AU  - Weiyu Sun
AU  - Tran Ngoc Huynh
AU  - Dawn Song
AU  - Bo Li
AU  - Ruoxi Jia
PY  - 2024
UR  - https://arxiv.org/abs/2406.17092
ID  - 2406.17092
ER  -