DREAM: Disentangling Risks to Enhance Safety Alignment in Multimodal Large Language Models

Add code
Apr 25, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: