最近一年的 AI 安全论文
本站共收录 71 篇- ALKN：面向持续 LLM 遗忘的知识否定自适应定位
Adaptive Localization of Knowledge Negation for Continual LLM Unlearning
- Decouple to Forget：面向大语言模型敏感信息遗忘的语义中介方法
Decouple to Forget: Semantic Mediation for Sensitive Unlearning in Large Language Models
- UnlearningSoup：大语言模型遗忘是否还需要反复调参？
UnlearningSoup: Is Repeated Tuning Necessary for Large Language Model Unlearning?
- TAME：用 SAE 抑制激活缓解 VLM 思维链混淆
Taming CoT Obfuscation in VLMs: From Mechanistic Evidence to Activation Enforcement
- 生成图像更易被遗忘：面向合成图像检测的机器遗忘视角
Generated Images Are Easier to Forget: A Machine Unlearning Perspective for Synthetic Image Detection
- DRNOISE：在误导性证据环境中评测深度研究智能体
DRNOISE: Benchmarking Deep Research Agents in Misleading Evidence Environments
- MLUBench：面向多模态大语言模型持续遗忘评测的基准
MLUBench: A Benchmark for Lifelong Unlearning Evaluation in MLLMs
- 超越同质性：迈向针对 GNN 的广义图重构攻击与防御
Beyond Homophily: Towards Generalized Graph Reconstruction Attack and Defense
- 可区分删除：在大语言模型遗忘中统一知识擦除与拒答
Distinguishable Deletion: Unifying Knowledge Erasure and Refusal for Large Language Model Unlearning
- 面向机器生成文本检测的多层级上下文 Token 关系建模
Multi-Level Contextual Token Relation Modeling for Machine-Generated Text Detection
- 当安全相冲突：用自适应安全引导解决文本到图像扩散模型的多类别有害冲突
When Safety Collides: Resolving Multi-Category Harmful Conflicts in Text-to-Image Diffusion via Adaptive Safety Guidance
- 超越原始检测分数：马尔可夫信息校准提升机器生成文本检测
Beyond Raw Detection Scores: Markov-Informed Calibration for Boosting Machine-Generated Text Detection
- AEGIS：面向扩散模型的对抗目标引导、无需保留数据的鲁棒概念擦除
AEGIS: Adversarial Target-Guided Retention-Data-Free Robust Concept Erasure from Diffusion Models
- 梯度上升真的必要吗？机器遗忘的“先记忆后遗忘”方法
Is Gradient Ascent Really Necessary? Memorize to Forget for Machine Unlearning
- ViKeR：面向多模态大语言模型遗忘的视觉引导关键 Token 正则化
Visual-Guided Key-Token Regularization for Multimodal Large Language Model Unlearning
- AdLift：把 2D 对抗扰动提升为 3D Gaussian Splatting 资产防护，抵御指令驱动编辑
AdLift: Lifting Adversarial Perturbations to Safeguard 3D Gaussian Splatting Assets Against Instruction-Driven Editing
署名以 dblp 和本人的 Google Scholar 为准；只列本站收录的 AI 安全论文。