Skip to content

Translation in progress Some entries still include text in the original language.

Oct 13 · Tuesday

All papers

Found 22 papers

Another 201papers do not have category tags yet and are not included in filtered results.

  1. CCDF dataset: Grok Imagine, Veo 3.1, Sora 2 forge surveillance footage, ten deepfake detectors basically cannot recognize it

    提出监控深伪基准 CCDF,在统一编码后评测检测器对伪造犯罪事故录像的识别

    Published
    Evaluated models
    C2P-CLIP, ForgeLens, D3-Im and others10

    Abstract提出消除格式捷径的监控深度伪造基准 CCDF,揭示现有检测器难以识别商用生成视频。

    Paper analysisRead analysis
  2. AttacksarXiv 2610.07323

    ATLAS:用自适应信任域与主动学习搜索潜在对抗样本集

    提出 ATLAS 用水平集估计恢复黑盒分类器的对抗输入集合

    Published
    Attacker
    Black-box
    Evaluated models
    LeNet5, ResNet50 (CIFAR-10 standard), ResNet50 (ImageNet standard) and others5
    AbstractATLAS 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。

    Paper analysisRead analysis
  3. DefensesarXiv 2610.02853

    研究提出基于收缩约束状态空间模型的有界可达性与越狱检测方法

    提出收缩约束 SSM 安全头,认证嵌入扰动并检测越狱

    Published
    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。。

    Paper analysisRead analysis
  4. DefensesarXiv 2610.00883

    DeBERTa-ConPara:面向部署的 AI 生成文本检测,推理端 Unicode 归一化提升对抗鲁棒性

    提出推理时 Unicode 归一化的检测器,识别遭字符级扰动的 AI 生成文本

    Published
    Evaluated models
    microsoft/deberta-v3-large, deberta-v3-base, bert-large-cased and others4
    Abstract推理时归一化消除两类 Unicode 攻击,训练时归一化则会删掉对抗监督。

    DeBERTa-ConPara 是一个面向部署的 AI 文本检测器:DeBERTa-v3-large 在多数据集原始文本上训练,推理前做确定性 Unicode 归一化,用一个固定阈值面对没有目标域标签的场景。

    Paper analysisRead analysis
  5. 研究:合成数据溯源识别在改写后准确率大幅下降

    检验合成数据来源识别是否等于再训练适配筛选

    Published
    Evaluated models
    Qwen2.5-3B-Instruct, DistilRoBERTa, word TF-IDF logistic regression and others5
    Abstract来源识别在原文上准确,改写后精确归因下降。

    作者在金融风险文本上把来源识别和训练样本是否有用拆成两次测量,并写明这不是电信部署实验。

    Paper analysisRead analysis
  6. 研究用 XAI 归因分析 Prompt Guard 2,同义词替换可翻转其提示注入判定

    用词级归因指导同义词替换,翻转提示注入检测并验证越狱

    Published
    Evaluated models
    Prompt Guard 2 (86M), Llama 3.1 8B

    Abstract作者称检测靠分散的词法线索。

    Paper analysisRead analysis
  7. AttacksarXiv 2609.15989

    Researchers propose plan injection attack that can bypass chain-of-thought monitoring

    提出 plan injection 攻击,植入良性伪装计划以绕过思维链监视器

    Published
    Attacker
    Gray-box
    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。

    Paper analysisRead analysis
  8. AttacksarXiv 2609.08236

    研究:内容不变的风格包装可翻转 LLM 安全评判器的判定

    提出内容不变样式包装,翻转安全评审器的判定

    Published
    Attacker
    Black-box
    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。

    Paper analysisRead analysis
  9. SinoGlyphBench:中文字形混淆下语言模型审核的诊断基准

    提出 SinoGlyphBench,用配对字形混淆诊断中文审核稳定性

    Published
    Evaluated models
    GPT-5.4 mini, GPT-5.5, DeepSeek V4 Flash and others12
    AbstractSinoGlyphBench 显示中文字形混淆会同时增加有害漏报和误报,全文扰动最重。

    SinoGlyphBench是一个配对诊断基准,用来看内容审核在可还原的中文字形混淆下是否还保持原决定。

    Paper analysisRead analysis
  10. 研究者提出最小范数攻击集成,用鲁棒性曲线替代单一 ε 评测

    提出最小范数攻击集成,用鲁棒性曲线评测图像分类器

    Published
    Evaluated models
    C1–C20(CIFAR-10,Table 1), I1–I10(ImageNet,Table 1)
    Abstract预算可控的最小范数集成用 AOI 逼近攻击前沿,DOI 给出不依赖单一ε的防御排序。

    这篇工作把对抗鲁棒评测从单一 ε、固定集成,改成可按查询预算收紧的曲线估计。

    Paper analysisRead analysis
  11. AttacksarXiv 2605.21541

    研究者提出 FRA-Attack 频域正则化方法,提升对闭源 MLLM 的迁移攻击效果

    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
    Attacker
    Black-box
    Evaluated models
    GPT-5.2, GPT-5.4, Claude-Opus-4.6 and others15
    AbstractFRA-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。

    Paper analysisRead analysis
  12. AttacksarXiv 2604.03117

    Unveiling Physical-World Semantic Vulnerabilities: Universal Adversarial Patches for Infrared Vision-Language Models

    Proposes the universal curved mesh patch UCGP, causing cross-task semantic bias in infrared vision-language models.

    Published
    Attacker
    Gray-box
    Evaluated models
    OpenAI CLIP ViT-L/14, OpenCLIP ViT-B/16, Meta-CLIP ViT-L/14 and others10
    AbstractUCGP 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.

    Paper analysisRead analysis
  13. AttacksarXiv 2603.21155

    Can LLMs fool graph learning? Exploring universal adversarial attacks on text-attributed graphs

    Propose BadGraph, jointly perturbing the topology and node text of text-attributed graphs under black-box settings

    Published
    Attacker
    Black-box
    Evaluated models
    DeepSeek-V3, Mistral-7B, Qwen-Plus and others10
    AbstractBadGraph 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.

    Paper analysisRead analysis
  14. AttacksarXiv 2602.09431

    Grounding-driven attack: Improving the transferability of encoder-based adversarial attacks to large vision-language models

    Propose GDA, which aligns image adversarial perturbations with text-grounded evidence and transfers attacks to vision-language models

    Published
    Evaluated models
    LLaVA, Qwen2.5-VL, InternVL3 and others10
    AbstractGDA aligns perturbations with text-grounded evidence, and the authors say this improves zero-query black-box transferability on heterogeneous LVLMs.

    GDA is a zero-query black-box encoder transfer attack for heterogeneous LVLMs.

    Paper analysisRead analysis
  15. AttacksarXiv 2602.00318

    Optimal transport-guided adversarial attacks against graph neural network bot detection

    提出 BOCLOAK 边编辑攻击,逃避图神经网络机器人检测

    Published
    Evaluated models
    BotRGCN, GAT, Simple-HGNN (S-HGN) and others5
    AbstractBOCLOAK 用邻域最优传输做受约束的 bot 逃避,并在三条社交图上比较检测器。

    BOCLOAK 是针对 GNN 社交 bot 检测器的测试时逃避方法,用来检查领域约束下的稳健性。

    Paper analysisRead analysis
  16. AttacksarXiv 2601.21531

    Adversarial robustness of large vision-language models under visual token compression

    提出 CAGE 攻击,在视觉 token 压缩下用图像扰动使预测偏离

    Published
    Attacker
    Gray-box
    Evaluated models
    LLaVA, Qwen2.5-VL, InternVL3-8B and others5
    AbstractCAGE 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。

    Paper analysisRead analysis