arXiv:2510.21885v1 Announce Type: cross Abstract: Large language models often lose previously aligned safety behaviors when fine-tuned on benign data, a phenomenon known as catastrophic forgetting. Prior work shows that adding random safety examples can mitigate this effect, but it remains unclear which examples are most effective. We propose a behavior-aware sampling framework that selects safety examples based on two complementary factors: instruction-response behavior (e.g., refusal versus compliance) and semantic diversity across harm categories. Systematic evaluation shows that this approach substantially reduces harmful outputs while maintaining helpfulness, achieving up to a 41% reduction in harmfulness with only 0.5% additional training data. These results highlight how targeted data selection can improve the safety and efficiency of fine-tuning at scale.