RSafe: Incentivizing proactive reasoning to build robust and adaptive LLM safeguards

Add code
Jun 09, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: