SafetyNet: Detecting Harmful Outputs in LLMs by Modeling and Monitoring Deceptive Behaviors

M. Chaudhary, F. Barez
May 2025
Type
Preprint
Publication
arXiv:2505.14300
Safety & Alignment