One-Shot Safety Alignment for Large Language Models via Optimal Dualization

Devs

One-Shot Safety Alignment for Large Language Models via Optimal Dualization | Read Paper on Bytez