Fetching the paper…

Virus: Harmful Fine-tuning Attack for Large Language Models Bypassing Guardrail Moderation · Around