Universal and Transferable Adversarial Attacks on Aligned Language Models

1 point by montenegrohugo 1 year ago | 0 comments