Uncovering Safety Risks in Open-source LLMs through Concept Activation Vector

Add code
Apr 18, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: