Jailbreak to Protect: Buffering and Reinforcing via Temporary Jailbreaking for Safe Fine-Tuning in Large Language Models 文章

ArXiv CS.CL2026-05-26NEWSen作者: Seokil Ham, Jaehyuk Jang, Wonjun Lee, Changick Kim
大语言模型
查看原文 →
关系图谱

Jailbreak to Protect: Buffering and Reinforcing via Temporary Jailbreaking for Safe Fine-Tuning in Large Language Models · 相关技术

相关技术

adapterODELLMlanguage model远程代码执行（RCE）safetyalignmentSAFEReferring expression comprehension (REC)Parts-of-Speech (POS) tagsNarrative Abstraction BenchmarkMECLoRALMMlarge language modelsHISForEffort Metric AttentionENADAPTARGANNICECLEAR