Spot Risks Before Speaking! Unraveling Safety Attention Heads in Large Vision-Language Models

Add code
Jan 03, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: