Google AI Introduces Consistency Training for Safer Language Models Under Sycophantic and Jailbreak Style Prompts
"Jailbreak Attacks" 相关文章
Learning to Detect Unknown Jailbreak Attacks in Large Vision-Language Models
Confusion is the Final Barrier: Rethinking Jailbreak Evaluation and Investigating the Real Misuse Threat of LLMs
Expanding our model safety bug bounty program
