1
Two to Tango: Coupled Task-Reference Selection for Safe LLM Fine-tuning
提出耦合任务与参考数据选择的新方法，保障大语言模型微调安全性。
arXiv:2606.09866v1 Announce Type: cross Abstract: Fine-tuning safety aligned large language models (LLMs) on downstream data improves adaptation but m…