TY  - RPRT
TI  - Combining Domain and Alignment Vectors to Achieve Better Knowledge-Safety Trade-offs in LLMs
AU  - Megh Thakkar
AU  - Quentin Fournier
AU  - Matthew Riemer
AU  - Pin-Yu Chen
AU  - Amal Zouaq
AU  - Payel Das
AU  - Sarath Chandar
PY  - 2025
UR  - https://arxiv.org/abs/2411.06824
ID  - 2411.06824
ER  -