Skip to content

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

Oct 13 · Tuesday

All papers

Found 3 papers
Filter5

Attack type

Jailbreaks8Prompt injection74Poisoning and backdoors16Extraction and theft3Detection evasion14Malicious use4

Attack surface / entry point

User input34Web pages and interfaces0Retrieval corpus3Inter-agent messages3Images, audio and video3Weights and gradients6

Layer

Model layer2Application layer1

System component

Multimodal3
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 203papers have no tags yet and are not included in filtered results.
Attack type：Extraction and theftAttack surface / entry point：Web pages and interfacesAttack surface / entry point：Images, audio and videoLayer：Model layerLayer：Application layerClear all

Another 203papers do not have category tags yet and are not included in filtered results.

  2. DefensesarXiv 2609.23586

    VidMark：面向视频生成模型知识产权保护的高效水印方案

    提出 VidMark 水印方案，核验视频生成模型的输出来源与所有权

    arXiv
    2609.23586
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Stable Video Diffusion (SVD), Open-Sora, Wan
    AbstractVidMark 加解码器引导微调，把模型签名写入 VGM 输出并用于两项核验。

    Huang 等人提出一套面向视频生成模型的生成内嵌水印方案，用来同时判断一条视频是否来自受保护模型，以及一个嫌疑模型是否为该模型的拷贝。

    Paper analysisRead analysis
  3. AttacksarXiv 2508.06837

    Prometheus：针对文生图扩散模型的免训练提示词窃取攻击

    提出 Prometheus，从展示图反推文生图模型的在售提示词

    arXiv
    2508.06837
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Attacker
    No-box (offline transfer)
    Evaluated models
    Midjourney, Leonardo.ai, DALL·E and others4
    Abstract训练无关的 proxy 搜索攻击。

    Prometheus 是针对文生图市场 showcase 的训练无关提示词窃取：攻击者看不到在售文本，只用本地开源 proxy 恢复 subject 和 modifier，并希望换 subject 后风格仍在。

    Paper analysisRead analysis
You have reached the end