Skip to content

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

Oct 12 · Monday

All papers

Found 10 papers
Filter5

Attack type

Jailbreaks10

Attack technique

Gradient and representation optimization15Automated search47Multi-turn escalation26Encoding and obfuscation40Role-play and persuasion20Conversation structure manipulation10Reasoning-chain manipulation7Cross-modal payloads29Defense-targeted10

Impact

Harmful content9

System type

Models and APIs9

Attacker access

Black-box7

Attack surface / entry point

User input10Web pages and interfaces0Tool and skill descriptions1

Layer

Prompt layer8
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 type：JailbreaksAttack technique：Conversation structure manipulationAttack surface / entry point：User inputAttack surface / entry point：Web pages and interfacesAttack surface / entry point：Tool and skill descriptionsClear all

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

  1. AttacksarXiv 2609.36956

    研究者提出 Controlled Decoding Attacks，通过纯文本接口越狱黑盒 LLM

    提出 BLINDBIAS，通过纯文本续写与前缀控制越狱黑盒模型

    arXiv
    2609.36956
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    GLM-5, Gemini-3.5-Flash, Qwen3-32B and others4
    Abstract作者认为，隐藏数值概率并不足以消除支持重复采样和前缀续写的解码攻击面。

    BLINDBIAS 是一种针对 sample-only 续写接口的 black-box 解码时越狱框架，研究没有目标数值概率时，采样能否支持逐步控制。

    Paper analysisRead analysis
  5. Analysis and theoryarXiv 2606.04168

    论文提出随机插入攻击，利用自回归一致性绕过 LLM 安全对齐

    解释自回归一致性如何导致浅层安全对齐，并提出随机插入攻击

    arXiv
    2606.04168
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Abstract作者用自回归一致性串起浅层对齐、插入攻击和最坏延续对齐。

    作者把浅层安全对齐和一类续写攻击都归到自回归一致性上。

    Paper analysisRead analysis
  6. InterpretabilityarXiv 2603.08234

    The Continuation vs. Refusal Debate: Mechanism Analysis of Continuation-Triggered Jailbreaks in Large Language Models

    Analyzing the attention head mechanism of continuation-triggered jailbreaks and proposing HCS

    arXiv
    2603.08234
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    LLaMA-2-7B-Chat, Qwen2.5-7B-Instruct
    AbstractThe authors state that jailbreaks arise from competition between two types of heads, and that HCS plus distillation can use this competition for inference-time intervention.

    The authors use attention head mechanism analysis for continuation-triggered jailbreaks: under the same harmful instruction, moving the continuation suffix from after the instruction to outside the prompt boundary makes the model more likely to continue generating along the suffix.

    Paper analysisRead analysis
  8. AttacksarXiv 2510.10271

    MetaBreak：通过特殊 token 操纵越狱在线 LLM 服务

    提出 MetaBreak，用特殊 token 操纵越狱在线 LLM 服务并绕过审核

    arXiv
    2510.10271
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    Llama-3.3-70B-Instruct, Qwen-2.5-72B-Instruct, Gemma-2-27B-Instruct and others10

    Abstract利用特殊 token 注入对抗平台模板包裹与审核，多场景验证了越狱有效性与互补性。

    Paper analysisRead analysis
  9. AttacksarXiv 2509.25624

    STAC：看似无害的工具调用链如何攻破 LLM Agent

    提出 STAC 框架，把恶意目标拆成多轮看似无害的工具调用链

    arXiv
    2509.25624
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Attacker
    Black-box
    Evaluated models
    GPT-4.1-2025-04-14, GPT-4.1-mini-2025-04-14, Qwen3-32B and others8

    Abstract作者提出并验证了智能体在序列化工具攻击下的普遍脆弱性，指出针对累积动作序列防御的必要性。

    Paper analysisRead analysis
You have reached the end