Fetching the paper…

ShieldLM: Empowering LLMs as Aligned, Customizable and Explainable Safety Detectors · Around