google/shieldgemma-2b

ShieldGemma-2b is a 2.6 billion parameter, decoder-only large language model developed by Google, built upon the Gemma 2 architecture. It is specifically designed for safety content moderation, targeting four harm categories: sexually explicit content, dangerous content, hate speech, and harassment. This model functions as a text-to-text classifier, outputting 'Yes' or 'No' to indicate policy violations, making it optimized for filtering user inputs and model outputs.