安全对齐虽关键，但一刀切的拒绝机制在网络安全等高风险领域适得其反，这篇论文揭示了系统性缺陷。
arXiv:2607.02714v1 Announce Type: cross Abstract: There is no doubt that safety alignment is an essential step in LLM training. However, conceptually …
从几何角度揭示安全对齐LLM的拒绝机制实为线性特征，提出对比对数引导方法验证其不稳定性。
arXiv:2606.22686v1 Announce Type: cross Abstract: Modern Large Language Models (LLMs) rely on extensive safety alignment, yet the mechanistic basis of…
用心理学原理优化大模型拒绝回答，减少生硬有害回应，提升AI交互安全性。
arXiv:2606.09697v1 Announce Type: new Abstract: Large language models (LLMs) routinely face requests that should be refused, creating a trade-off betw…
提出首个专注于AI Agent网络安全拒绝机制的框架，弥补现有基准只重能力、忽视安全限制的空白。
arXiv:2606.02644v1 Announce Type: cross Abstract: Agentic scaffolds have dramatically improved LLM performance on complex, long-horizon tasks, yieldin…
揭秘链式思维推理如何打破AI拒绝行为的方向性操控，大模型安全新视角
arXiv:2605.26772v1 Announce Type: new Abstract: Large reasoning models (LRMs) generate chain-of-thought (CoT) traces before producing final outputs, i…