Skip to content

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

Oct 12 · Monday

All papers

Found 8 papers
Filter4

Attack type

Extraction and theft5

Impact

Harmful content9Unauthorized actions3Data and privacy leaks8Output manipulation9Availability and reliability3

System type

Models and APIs8

Attacker access

White-box3

Attack surface / entry point

User input9Code repositories and dependencies0Training data5Weights and gradients8

Layer

Model layer4Application layer3Training and data layer8

System component

Fine-tuning and open weights3
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 233papers have no tags yet and are not included in filtered results.
Impact：Data and privacy leaksAttack surface / entry point：Code repositories and dependenciesAttack surface / entry point：Weights and gradientsLayer：Training and data layerClear all

Another 233papers do not have category tags yet and are not included in filtered results.

  1. AttacksarXiv 2610.05453

    研究者提出 Privacy-Leaking Watermarks，可在本地多模态模型中泄露对话隐私

    提出 PLW，用微调把先前对话敏感触发编码进后续生成图像

    arXiv
    2610.05453
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    OmniGen2, BAGEL-7B

    AbstractPLW 把先前对话线索写成后续图像中的隐水印，检出随触发、模型和间隔变化。

    Paper analysisRead analysis
  2. AttacksarXiv 2610.04128

    研究显示拆分学习中的梯度可提升客户端文本重建率

    衡量分裂学习中梯度相对激活对客户文本重建的额外贡献

    arXiv
    2610.04128
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    GPT-2 (openai-community/gpt2, 124M), Qwen3-0.6B
    Abstract激活已恢复大部分 token。

    这篇工作测量 split learning 分裂点上的文本泄漏：激活单独能恢复多少，训练回传的梯度再增加多少，以及按 token 还是按整篇计分会不会改写结论。

    Paper analysisRead analysis
  3. AttacksarXiv 2610.01365

    ReGap：无需真实隐私数据即可通过微调重新提取模型隐私关联

    提出 ReGap，用自生成伪监督微调恢复模型中的隐私关联

    arXiv
    2610.01365
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Attacker
    White-box, Gray-box
    Evaluated models
    GPT-2 Small, GPT-2 Medium, GPT-2 Large and others6

    Abstract提出无真实私有监督的恢复攻击 ReGap，实验显示微调可通过自生成数据唤醒模型潜在隐私关联，突显后适应隐私风险。

    Paper analysisRead analysis
  6. DefensesarXiv 2609.20532

    面向旧款 GPU 的可验证差分隐私训练框架：用 CPU 侧 TEE 校验 GPU 梯度

    AI summary论文提出一种在旧款 GPU 上实现可验证差分隐私（DP）训练的框架，用 CPU 侧 TEE 配合不受信任的 GPU 完成校验。

    arXiv
    2609.20532
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    5-layer MLP, WideResNet, GPT-2-medium and others4
    AbstractCPU TEE 概率抽查不可信 GPU 上的 DP-SGD，以小开销约束频繁偏离。

    TEE-Certified DP 用已部署的 CPU TEE 监视不可信旧 GPU 上的 DP-SGD，使监管方能在不看训练数据的情况下检查声明的隐私机制是否被执行。

    Paper analysisRead analysis
  7. DefensesarXiv 2609.15671

    QPriv-VL：面向视觉语言模型的问题引导 token 裁剪隐私防御

    提出 QPriv-VL，用问题引导剪枝降低分布式视觉问答的隐私泄露

    arXiv
    2609.15671
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    LLaVA-1.5-7B, Llama-2
    AbstractQPriv-VL 在切层前按问题相关度与 DINOv2 敏感度剪视觉 token，作者称能在约四成 token 下兼顾准确率与四类攻击。

    QPriv-VL 把问题引导的视觉 token 剪枝做成分布式 VQA 的客户端隐私防御，而不是只做集中式推理加速。

    Paper analysisRead analysis
  8. AttacksarXiv 2609.04382

    研究揭示 split-LLM 训练中返回梯度泄露真实行，前向隐私门仍通过

    揭示拆分训练返回梯度以零支撑暴露诱饵行并泄露训练内容

    arXiv
    2609.04382
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Attacker
    White-box
    Evaluated models
    Qwen3-0.6B, 35B-A3B
    Abstract论文揭示拆分训练中反向梯度零支撑完全暴露诱饵行，联合视图突破门控，裁剪加噪可有效缓解。

    这是一项关于双节点大语言模型拆分训练系统的系统安全案例研究，揭示了评估工具遗漏反向信道所导致的静默隐私失效。

    Paper analysisRead analysis
You have reached the end