Skip to content

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

Oct 13 · Tuesday

Jailbreaks and attacks · Papers

Found 5 papers
Filter3

Attack type

Jailbreaks5

Attack technique

Gradient and representation optimization5Automated search16Multi-turn escalation5Encoding and obfuscation18Cross-modal payloads33

Impact

Harmful content4

System type

Models and APIs4

Attacker access

Black-box3White-box3

Attack surface / entry point

Images, audio and video5

Layer

Prompt layer4

System component

Memory0Vision3Multimodal5
Select any value within a category; different categories must all match. Counts include the other filters you have selected.
Attack technique：Gradient and representation optimizationSystem component：MemorySystem component：MultimodalTopic: Jailbreaks and attacksClear all
  3. AttacksarXiv 2605.07399

    研究者提出 GPO-V，用全局概率优化越狱扩散视觉语言模型

    提出 GPO-V，用全局概率优化的视觉扰动越狱扩散视觉语言模型

    arXiv
    2605.07399
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    White-box, Black-box
    Evaluated models
    LLaDA-V, LaViDA
    Abstract作者用视觉扰动改 dVLM 的全局生成概率，以绕过两类拒绝。

    GPO-V 把越狱从“开头必须顺从”改成“整段去噪概率朝肯定方向走”。。

    Paper analysisRead analysis
  5. AttacksarXiv 2511.16110

    MFA 框架在 17 个视觉语言模型上取得 58.5% 攻击成功率

    提出 MFA 框架，协同突破视觉语言模型的对齐、提示与审核层

    arXiv
    2511.16110
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    MiniGPT-4, LLaMA-4-Scout-Inst, LLaMA-3.2-11B-Vision-Instruct and others17

    Abstract作者称多层防御的 VLM 仍存在跨模型漏洞，MFA 整合三项技术实现了端到端越狱。

    Paper analysisRead analysis
You have reached the end