Explore definitions and dynamic coverage analytics for the core concepts shaping artificial intelligence.
Jailbreaking is a subset of adversarial prompt attacks where users structure prompt commands to bypass the safety alignment, moral rules, and system filters of Large Language Models.
