UofTCSSLab/SIREN-Qwen3-0.6B
Updated • 4
SIREN
SIREN is a lightweight guard model that detects harmful content from LLM internal representations.
SIREN is a lightweight guard model that detects harmful content from LLM internal representations.