Skip to content

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

Oct 12 · Monday

All papers

Found 5 papers
Filter4

Attack technique

Gradient and representation optimization7Automated search27Multi-turn escalation4Encoding and obfuscation21Role-play and persuasion5Reasoning-chain manipulation4Cross-modal payloads10Defense-targeted18Dormant triggers29

Impact

Output manipulation3

System type

Models and APIs3

Attack surface / entry point

User input3

Layer

Prompt layer3

System component

Memory0Guardrails and judges5Embeddings0Multimodal3
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.
Attack technique：Role-play and persuasionSystem component：MemorySystem component：Guardrails and judgesSystem component：EmbeddingsClear all

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

  1. AttacksarXiv 2609.08236

    研究：内容不变的风格包装可翻转 LLM 安全评判器的判定

    提出内容不变样式包装，翻转安全评审器的判定

    arXiv
    2609.08236
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    GPT-4o-mini (gpt4o), GPT-4o-mini (strongreject), DeepSeek-Chat and others7
    Abstract论文揭示特定安全评审器易受保持内容不变的外层包装诱导改变判定，并指出安全评估应将判定稳定性作为重要考量指标。

    这是一项关于大语言模型自动化安全评审系统表面样式脆弱性的评估研究。

    Paper analysisRead analysis
  2. Evaluation and benchmarksarXiv 2606.02630

    MultiTurnPSB

    构建 MultiTurnPSB，评测四轮医疗越狱与分类器输入防御

    arXiv
    2606.02630
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    GPT-4.1-mini, Claude Sonnet 4.5
    Abstract多轮攻击加剧医疗模型风险，分类器干预有效但受高误报率制约。

    MultiTurnPSB 针对患者端医疗大模型在持续交互中的安全风险，构建了包含三种攻击模态的四轮对抗评测基准。

    Paper analysisRead analysis
  3. Evaluation and benchmarksarXiv 2604.01444

    Cooking Risks: Benchmarking and Mitigating Food Safety Risks in Large Language Models

    Propose FoodGuardBench and the guardrail FoodGuard-4B to evaluate and reduce food hazard jailbreaks

    arXiv
    2604.01444
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    Claude-3.7-Sonnet, GPT-4o, GPT-4.1 and others10
    Paper analysisRead analysis
  4. AttacksarXiv 2601.21360

    Compliance Paradox: Semantic-Instruction Decoupling in Automated Academic Code Evaluation

    提出 AST-ASIP ，把对抗指令植入编译器忽略区域以劫持代码评分

    arXiv
    2601.21360
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    DeepSeek-V3.2, Qwen3-235B, GPT-OSS-120B and others9
    Abstract作者称 RLHF 的服从性使评分脱离代码。

    这篇工作把自动代码评分中的失败称为 Compliance Paradox：评分模型可以离开提交逻辑，转而执行嵌在可编译代码里的隐藏指令。

    Paper analysisRead analysis
You have reached the end