Skip to content

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

Oct 13 · Tuesday

All papers

Found 10 papers
Filter4

Attack type

Jailbreaks50Prompt injection56Poisoning and backdoors40Extraction and theft10Detection evasion18Malicious use11

Impact

Data and privacy leaks10

System type

LLM applications8

Attacker access

Black-box5

Attack surface / entry point

User input7

Layer

Application layer8

System component

MCP and skills3RAG8Monitors0Reasoning chains5Fine-tuning and open weights4Vision4Audio3Image generation5Embeddings5Multimodal12
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 213papers have no tags yet and are not included in filtered results.
Attack type：Extraction and theftSystem component：RAGSystem component：MonitorsSystem component：VisionClear all

Another 213papers do not have category tags yet and are not included in filtered results.

  3. Evaluation and benchmarksarXiv 2609.33481

    IDUnlearn-Bench：面向视觉语言模型的个体级遗忘基准

    提出 IDUnlearn-Bench，评测 VLM 个体级遗忘后的身份残留

    arXiv
    2609.33481
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    Qwen3-VL-2B, Qwen3-VL-4B, Qwen3-VL-8B and others6
    AbstractIDUnlearn-Bench 用四族探针显示：压低直接属性访问后，身份仍常能被反向查出、绑定或重构。

    IDUnlearn-Bench 评测 VLM 是否忘掉一个可跨模态识别的人，而不只是答不出某条监督事实。

    Paper analysisRead analysis
  4. 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
  5. AttacksarXiv 2609.14649

    CIG-MIA：利用上下文信息增益对 RAG 系统发起成员推断攻击

    提出 CIG-MIA，用上下文信息增益推断 RAG 知识库成员身份

    arXiv
    2609.14649
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Attacker
    Gray-box, Black-box
    Evaluated models
    Llama-3.1-8B-Instruct, Mistral-7B-Instruct-v0.3, Llama-2-13b-chat-hf and others6
    AbstractCIG-MIA 用注入前后的答案似然差推断 RAG 库成员，灰盒可分性高于所列基线。

    CIG-MIA 是针对 RAG 外部知识库的成员推断攻击，同时覆盖灰盒和纯文本黑盒。

    Paper analysisRead analysis
  6. DefensesarXiv 2609.04767

    SHAQ: Privacy-preserving retrieval in vector databases using shadow queries

    Proposes SHAQ, which replaces document embeddings with shadow query embeddings to defend against embedding inversion.

    arXiv
    2609.04767
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Attacker
    Black-box
    Evaluated models
    GTR-T5 encoder, DPR
    AbstractSHAQ replaces document embeddings with shadow query embeddings.

    SHAQ is a defense against EIA for cloud vector databases: instead of directly modifying document embeddings, it replaces them with shadow query embeddings.

    Paper analysisRead analysis
  9. DefensesarXiv 2605.19847

    论文揭示多租户 RAG 账号合谋使隐私泄露随账号数线性增长

    提出零知识审计协议，核验多租户 RAG 的合谋差分隐私

    arXiv
    2605.19847
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Abstract揭示多租户 RAG 合谋泄露按根号 k 放大，提出带零知识证明的无泄露审计协议。

    论文针对多租户 RAG 服务在多账户协同攻击下隐私边界失效的问题展开研究。

    Paper analysisRead analysis
  10. Evaluation and benchmarksarXiv 2602.09319

    Benchmarking Knowledge Extraction Attacks and Defenses in RAG

    建立 RAG 知识抽取攻防基准，统一评测抽取攻击与防御

    arXiv
    2602.09319
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    GPT-4o, GPT-4o-mini, Llama3-8B-Instruct and others8
    Abstract统一基准比较 RAG 知识抽取攻防，作者称检索与生成都要优化，单一防御不能完全防护。

    作者给出一个比较 RAG 知识抽取攻击与防御的统一基准。

    Paper analysisRead analysis
You have reached the end