Translation in progress Some entries still include text in the original language.
Oct 12 · Monday
All papers Found 18 papers Another 233 papers do not have category tags yet and are not included in filtered results.
“
About this collection Collection in progress We are building a large AI safety paper collection to help researchers find work related to their interests. Papers are collected automatically andfiltered only for relevance to AI safety , without judging quality, novelty or reading value. The default “Annotated first” shows papers with tags or published paper analyses first. Each group is ordered by original publication date, newest first, with undated papers last. “Newest first” uses publication dates only. This order is not a quality ranking.
One-line summaries, tags and paper analyses are prepared by AI from the papers. Language models may have systematic biases and produce hallucinations.We plan to carry out human annotation and review to reduce errors in this content.
提出 ATLAS 用水平集估计恢复黑盒分类器的对抗输入集合
Published Oct 6
Evaluated models LeNet5 , ResNet50 (CIFAR-10 standard) , ResNet50 (ImageNet standard) and others5 Abstract ATLAS 用主动水平集估计构造黑盒对抗集，作者称其集合比单点攻击更具代表性。ATLAS 是一个查询式 black-box 框架，用来构造诱发失败的对抗输入集合，供鲁棒性审计使用。
提出收缩约束 SSM 安全头，认证嵌入扰动并检测越狱
Published Oct 2
Evaluated models ToySSMClassifier , S4 safety head Abstract 收缩给出玩具 LTI 的有界认证。给 SSM 安全头一个收缩条件，并在越狱检测上使用收缩正则的 S4 头。 。
提出 UNTAG，绕过开源权重模型的 trigger-tag 滥用检测
Published Oct 2
Evaluated models Llama2 (7B) , Llama3 (8B) , Qwen2.5 (7B) Abstract 作者用 UNTAG 评估 trigger-tag，认为其不能充当开放权重滥用检测的主要边界。UNTAG 是针对开放权重 LLM 上条件滥用检测 trigger-tag 的统一攻击评估：信号可以在解码期注入，也可以写进权重，但攻击者若能改输出或改已发布权重，检测可以失效。
提出 Fed-ADR，用编排服务器协同操纵梯度攻破联邦学习并检测恢复
Published Sep 23
Evaluated models CNN（MNIST：2 Conv + 3 FC） , CNN（Fashion-MNIST：6 Conv + 1 FC） , ResNet18（CIFAR-10） Abstract Fed-ADR 用 gray-box 编排攻击绕过多种聚合防御，再用 Hessian 一致性检测和窗口内恢复把准确率拉回。Fed-ADR 是一套面向联邦学习的攻击、检测与恢复框架：编排服务器改写合谋客户端的梯度，使参数服务器的现有防御把投毒更新当成良性，同时在发现后用历史梯度把全局模型原地修回来。
提出 BoundStyle 与 StyleAT，对抗训练防御人脸识别语义攻击
Published Sep 20
Evaluated models MobileFaceNet (MobileFace) , ResNet (ResNet-152, IR-SE) , RepVGG and others7 Abstract BoundStyle 用 StyleGAN3 有界潜编辑快速攻击 FR。StyleAT 用可调的 StyleGAN3 潜空间攻击 BoundStyle 对人脸识别做对抗训练，以抵抗一般语义编辑，并在评测中面对训练时未见的 DiffPrivate。
提出 plan injection 攻击，植入良性伪装计划以绕过思维链监视器
Published Sep 15
Evaluated models Qwen2.5-Coder-7B-Instruct , Qwen3-8B , DeepSeek-R1-0528 and others6 Abstract 论文指出模型会将外部上下文中的伪装计划改写为自身推理，使思维链监视器失效甚至反向合理化恶意行为。
复现 Tree-Ring 语义水印的黑盒伪造并恢复检测统计量
Published Sep 11
Evaluated models Stable Diffusion XL base 1.0 提出内容不变样式包装，翻转安全评审器的判定
Published Sep 8
Evaluated models GPT-4o-mini (gpt4o) , GPT-4o-mini (strongreject) , DeepSeek-Chat and others7 Abstract 论文揭示特定安全评审器易受保持内容不变的外层包装诱导改变判定，并指出安全评估应将判定稳定性作为重要考量指标。这是一项关于大语言模型自动化安全评审系统表面样式脆弱性的评估研究。
提出黑盒攻击 DRIFT，偏转生成轨迹以去除扩散水印
Published Sep 8 提出 SPARED，用对抗编辑数据训练基于推理的 AI 生成图像检测器
Published Aug 13
Evaluated models Qwen3.5-9B Abstract SPARED 用防捷径的编辑对抗环训练 9B 推理检测器。SPARED 是一套用于可解释 AI 生成图像检测的攻防强化学习流程，报告模型是 Qwen3.5-9B 上的第三轮防御器。
提出 ColluSkill，用跨技能组合绕过 Agent 技能扫描器
Published Aug 10
Evaluated models GPT-5.5 , DeepSeek-V4-Pro , GLM-5.2 Abstract 论文揭示了单技能孤立扫描的盲区，提出多技能链攻击 ColluSkill 与上下文防御 ChainGuard 并验证了有效性。
提出 FRA-Attack，用频域对齐与梯度正则做可迁移定向视觉攻击
Published May 20
Evaluated models GPT-5.2 , GPT-5.4 , Claude-Opus-4.6 and others15 Abstract FRA-Attack 以频域对齐和梯度低通做可迁移定向攻击，闭源面板上作者称整体更强，Gemma 除外。FRA-Attack 是面向闭源 MLLM 的迁移式定向图像攻击。作者要解决的是：空间域局部对齐重复全局低频，代理梯度的高频成分又把更新拉向代理自身。
Proposes the universal curved mesh patch UCGP, causing cross-task semantic bias in infrared vision-language models.
Published Apr 3
Evaluated models OpenAI CLIP ViT-L/14 , OpenCLIP ViT-B/16 , Meta-CLIP ViT-L/14 and others10 Abstract UCGP uses one shared curved-grid patch to simultaneously lower the classification and open-ended outputs of IR-VLM.UCGP is a universal physical adversarial patch for infrared vision-language models. The question is: can a patch that can be attached and reused across samples cause stable semantic deviations in the infrared visual evidence shared by classification, description, and VQA. The patch is constrained to a low-frequency curved mesh.
Propose BadGraph, jointly perturbing the topology and node text of text-attributed graphs under black-box settings
Published Mar 22
Evaluated models DeepSeek-V3 , Mistral-7B , Qwen-Plus and others10 Abstract BadGraph uses LLMs to jointly perturb TAG topology and text under black-box settings, reducing accuracy across multiple reasoners.BadGraph is a black-box universal attack for node classification on text-attributed graphs, used to check whether the same topology plus text perturbation can drive both GNN reasoners and LLM reasoners toward predetermined wrong labels.
提出 CAGE 攻击，在视觉 token 压缩下用图像扰动使预测偏离
Published Jan 29
Evaluated models LLaVA , Qwen2.5-VL , InternVL3-8B and others5 Abstract CAGE 对齐压缩瓶颈后，压缩 LVLM 的稳健准确率低于全 token 编码器攻击。CAGE 研究的是：LVLM 在推理时剪枝或合并视觉 token 之后，对抗稳健性该怎么估。
Attacks arXiv 2510.10987 · Oct 13, 2025 提出 DITTO 框架，蒸馏提取水印信号并注入外部模型以伪造归属
Published Oct 13, 2025
Evaluated models Llama3.1-8B , Llama3.2-3B , GPT-OSS 20B and others4 Abstract 论文针对大模型水印归属假设的脆弱性，提出黑盒仿冒攻击 DITTO，实现跨架构和采样方案的高检测率与低困惑度仿冒。
Attacks arXiv 2509.09112 · Sep 11, 2025 提出字符级扰动与遗传算法去除 LLM 水印
Published Sep 11, 2025
Evaluated models OPT-1.3B , LLaMA-3-8B Abstract 字符级扰动靠更大攻击范围移除水印。这篇工作评估推理时 LLM 水印在黑盒、有限查询下的移除难度，并比较字符级、token 级和句子级编辑。
Attacks arXiv 2503.02174 · Mar 4, 2025 提出 AdvTok 对抗分词攻击，不改文本只改分词来越狱并规避安全分类器
Published Mar 4, 2025
Evaluated models Llama3.2-1B , Llama3.2-3B , Llama3.1-8B and others7 Abstract 作者称非规范分词能躲开对齐并仍得到针对原请求的回复。作者用非规范分词攻击已对齐的子词 LLM：恶意字符串的字符不变，只改 token 边界。
You have reached the end