Paper recorded by Signals 4 on 2026-10-27 in cs.CL. Abstract reproduced from arXiv; link to the original below.
Published 2026-10-27 on arXiv · recorded by Signals 4 on 2026-10-28
Category: cs.CL · 自然语言处理 · first seen 2026-10-28
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 that identifies fine-tuning examples whose gradients align with unsafe behavior