New Framework Developed to Combat Jailbreak Attacks on AI Models
Researchers have introduced the sysname Framework, designed to protect large language models from jailbreak attacks by addressing specific vulnerabilities and reducing the Attack Success Rate.