Skip to content

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

Oct 12 · Monday

All papers

Found 7 papers
Filter5

Attack type

Jailbreaks7Prompt injection17Poisoning and backdoors34Detection evasion32Malicious use3

Impact

Harmful content202Unauthorized actions3Output manipulation7

System type

Models and APIs7coding agent0Browser and computer use0

Attack surface / entry point

User input5

Layer

Prompt layer4Model layer3

System component

Multimodal3
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 240papers have no tags yet and are not included in filtered results.
Attack type：JailbreaksImpact：Output manipulationSystem type：Models and APIsSystem type：coding agentSystem type：Browser and computer useClear all

Another 240papers do not have category tags yet and are not included in filtered results.

  1. Evaluation and benchmarksarXiv 2610.11112

    EpiReal-Bench：商用图像生成模型虚假声明红队基准

    构建 EpiReal-Bench，评测图像生成器将虚假主张渲染为可信证据

    arXiv
    2610.11112
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    GPT-Image-2, Nano Banana 2, Grok Imagen 2.0 and others4
    Abstract四款商用图像模型能把假主张画成可信证据。

    EpiReal-Bench 用来量商用图像生成器会不会把虚假现实主张画成可信视觉证据。

    Paper analysisRead analysis
  2. Analysis and theoryarXiv 2609.36477

    研究发现护栏模型在基座模型不确定处过度自信

    诊断护栏相对基座在对抗提示上的置信失校

    arXiv
    2609.36477
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Llama-Guard 7B, Llama-Guard-2 8B, Llama-Guard-3 8B and others10
    Abstract对抗包装使 guard 高置信漏检，而对应 base 在同一输入上常更不确定。

    诊断性分析：五个提示词分类 guard 的置信度，在清洁输入上可以较校准，在越狱包装后则变成高置信度的错误判定。

    Paper analysisRead analysis
  5. AttacksarXiv 2609.03247

    研究揭示大语言模型的自发身份认证失败：五款模型的分阶段开发者身份实验

    提出 MIPC 与 CFA，观察模型自拟测验后接受开发者身份

    arXiv
    2609.03247
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    ChatGPT, Claude, Qwen3.7-Plus and others5
    Abstract自拟测验可造成对话层虚假认证，但所测授权边界并未因此改变。

    这项工作把“模型自己出题、自己批改、再宣布身份已核验”定义成对话层认证失败，并与后端提权分开。

    Paper analysisRead analysis
  7. Evaluation and benchmarksarXiv 2602.19450

    对面向可信执行环境的 Claude Opus 与 ChatGPT 安全顾问进行红队测试

    用 TEE-REDBENCH 评测 TEE 顾问的机制幻觉与提示诱导失败

    arXiv
    2602.19450
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    ChatGPT-5.2, Claude Opus-4.6
    Abstract作者用 TEE-REDBENCH 测两款顾问，称失败可迁移且流水线控制可降低发生率。

    TEE-REDBENCH 红队评测被当作 Intel SGX 与 Arm TrustZone 安全顾问的 ChatGPT-5.2 和 Claude Opus-4.6，看架构推理、远程证明语义、缓解建议和策略约束下的拒答，并测提示诱导失败能否跨助手迁移。作者把助手看成随机策略。

    Paper analysisRead analysis
You have reached the end