TY  - RPRT
TI  - Safety Alignment Backfires: Preventing the Re-emergence of Suppressed Concepts in Fine-tuned Text-to-Image Diffusion Models
AU  - Sanghyun Kim
AU  - Moonseok Choi
AU  - Jinwoo Shin
AU  - Juho Lee
PY  - 2024
UR  - https://arxiv.org/abs/2412.00357
ID  - 2412.00357
ER  -