Follow the Gradient: Tracing Safety Degradation Back to Training Data
Fine-tuning Large Language Models (LLMs) on instruction data can quietly erode their safety, even when no training example looks harmful. Because this safety-degrading data looks benign, standard content filters struggle to catch it. We tackle this problem by developing SimpleLoRIF, a scalable gradient embedding method...
Saket Reddy, Amro Abbas, Anshuman Suri et al.
· 0 citations