大模型安全：从对齐问题到对抗性攻击的深度分析
"Adversarial Attacks" 相关文章
大模型中毒记
NeurIPS2025 | 攻破闭源多模态大模型：一种基于特征最优对齐的新型对抗攻击方法
Enhancing Genomic Foundation Model Robustness through Iterative Black-Box Adversarial Training
Enhancing Genomic Foundation Model Robustness through Iterative Black-Box Adversarial Training
Vision Transformers: the threat of realistic adversarial patches
Enhancing NLP Models for Robustness Against Adversarial Attacks: Techniques and Applications
EMNLP 2025 | LLM也会“装成人”？对比重写CoPA挑战AI文本检测可信度
