Latent-space adversarial training with post-aware calibration for defending large language models against jailbreak attacks

Add code
Jan 18, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: