Skip to content

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

Oct 12 · Monday

All papers

Found 11 papers
Filter4

Attack type

Extraction and theft7

Impact

Harmful content11Unauthorized actions2Data and privacy leaks9Output manipulation16Availability and reliability3

System type

Models and APIs5LLM applications4

Attacker access

Black-box4

Attack surface / entry point

Retrieval corpus3Images, audio and video3

Layer

Model layer4Application layer6

System component

MCP and skills59Memory39RAG11Guardrails and judges29Monitors30Reasoning chains10Permissions and sandboxing69Fine-tuning and open weights12Vision20Audio6Image generation5Embeddings5Multimodal30
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 232papers have no tags yet and are not included in filtered results.
Impact：Unauthorized actionsImpact：Data and privacy leaksSystem component：AudioSystem component：EmbeddingsClear all

Another 232papers do not have category tags yet and are not included in filtered results.

  3. Evaluation and benchmarksarXiv 2609.35922

    VoxParity 评测：语音智能体在需要听声判断时仍按文字行事

    提出 VoxParity，固定转写只改音频，评测语音智能体是否按行业规则改工具调用

    arXiv
    2609.35922
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    MiMo-V2.6-Pro, gemini-3.7-flash, Qwen3.8-Omni and others15
    AbstractVoxParity 用固定转写的音频对比和 words-only null，测量语音智能体是否按所听内容改变工具调用。

    VoxParity 测量语音智能体在词语不变、只有声音改变时，是否按行业规则改它所执行的工具调用。

    Paper analysisRead analysis
  5. DefensesarXiv 2609.27399

    GUARD：在 CosyVoice2 上做说话人身份遗忘以阻止重识别

    提出 GUARD，在零样本 TTS 上用激活转向遗忘说话人身份以阻止再识别

    arXiv
    2609.27399
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    CosyVoice2-0.5B, F5-TTS, FireRedTTS
    AbstractGUARD 用门控激活转向把退出说话人推向冒充相似度，以抑制再识别并保住音质。

    GUARD 用门控激活转向做说话人身份遗忘，使零样本 TTS 对退出说话人难以再识别。

    Paper analysisRead analysis
  6. Evaluation and benchmarksarXiv 2609.09404

    MMPIBench：多模态提示注入对六种 Agent 框架的跨框架评测

    用 MMPIBench 评测多模态提示注入对 Agent 框架的影响

    arXiv
    2609.09404
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    Claude Opus 4.8, GPT-5.4, Gemini 3.1 Pro and others6

    Abstract论文提出 MMPIBench 评测多模态提示注入对智能体的影响，发现模型主导行为、尝试远超完成，音频通道缺乏防护。

    Paper analysisRead analysis
  8. 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. AttacksarXiv 2609.01723

    研究者提出针对微调 TTS 模型的黑盒成员推断攻击

    提出针对微调 TTS 的黑盒成员推断攻击

    arXiv
    2609.01723
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    CosyVoice2, F5-TTS, XTTS-v2
    Abstract黑盒 TTS 成员推断以 recitation 加分层语音表示为分数，说话人级强于记录级，DP-SGD 可压到随机附近。

    作者给出一个针对微调文本转语音模型的黑盒成员推断框架，分别审计说话人是否进入微调集，以及某一条文本–音频记录是否进入微调集。微调 TTS 以合成文本和参考语音双条件生成，既有生成模型 MIA 的单条件查询和图像/文本上的逐点比较不能直接套用。作者用 scorable extent 与 memorization elicitation 比较五种查询，认为只有 recitation（用目标转写和干净目标语音要求模型复现）同时满足两条，并把它作为主查询。

    Paper analysisRead analysis
  10. AttacksarXiv 2607.01276

    研究者提出嵌入推断攻击 EIA，可在黑盒 IR 系统中识别所用嵌入模型

    提出嵌入推断攻击 EIA，在黑盒 IR 中识别所用嵌入模型

    arXiv
    2607.01276
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Attacker
    Black-box
    Evaluated models
    e5-small, e5-small-v2, e5-base and others19
    AbstractEIA 用无序文档集的差异从候选中识别嵌入模型，并给出 reranker 与阈值下的计数。

    EIA 是在黑盒 IR 上识别所用嵌入模型的攻击。。

    Paper analysisRead analysis
  11. AttacksarXiv 2602.01757

    Zero2Text: Training-Free Cross-Domain Text Embedding Inversion Attack

    提出 Zero2Text，无需训练地跨域反演文本嵌入

    arXiv
    2602.01757
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    GTR-Base, Qwen3-Embedding-0.6B, Text-Embedding-3-small and others4
    Abstract零训练在线对齐可在未知域、无泄漏对时从黑盒 embedding 还原句子。

    Zero2Text 是一种不训练 decoder、也不使用泄漏文本–向量对的文本 embedding 反演，针对 RAG 向量库在严格黑盒、目标域未知时的还原。

    Paper analysisRead analysis
You have reached the end