Browsing: Shielding

This paper provides a framework to represent, quantify, and evaluate safety in AI. It uses a logic-based approach rather than a numerical one to…