"RAJ-PGA: Reasoning-Activated Jailbreak and Principle-Guided Alignment Framework for Large Reasoning Models"
收藏DataCite Commons2026-01-15 更新2026-05-03 收录
下载链接：
https://ieee-dataport.org/documents/raj-pga-reasoning-activated-jailbreak-and-principle-guided-alignment-framework-large
下载链接
链接失效反馈官方服务：
资源简介：
"The PGA dataset is a safety alignment dataset specifically designed for Large Reasoning Models (LRMs). It contains 3,989 rigorously verified high-quality samples, aimed at addressing the unique Reasoning-Activated Jailbreak problem in LRMs. Unlike traditional Simple Refusal datasets, PGA strives to shift the alignment paradigm toward Intrinsic Safe Reasoning, thereby enhancing model safety while preserving deep reasoning capabilities."
提供机构：
IEEE DataPort
创建时间：
2026-01-15
