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

Poisoning and backdoors6

Attack technique

Gradient and representation optimization19Automated search52Multi-turn escalation24Encoding and obfuscation44Role-play and persuasion24Conversation structure manipulation11Reasoning-chain manipulation11Cross-modal payloads28Defense-targeted13Dormant triggers7

Impact

Output manipulation3

System type

Models and APIs5AI Agent10coding agent2Embodied systems and robotics0

Attack surface / entry point

User input7Tool and skill descriptions3Images, audio and video8Training data15Weights and gradients14Distribution channels14

Layer

Training and data layer3

System component

Fine-tuning and open weights3
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 233papers have no tags yet and are not included in filtered results.
Attack technique：Dormant triggersSystem type：Models and APIsSystem type：coding agentSystem type：Embodied systems and roboticsAttack surface / entry point：User inputClear all

Another 233papers do not have category tags yet and are not included in filtered results.

  1. AttacksarXiv 2610.12644

    Cancerous backdoor attack: progressive activation via invisible tokens and a stage controller.

    Proposes the Cancerous backdoor, which progressively corrupts reasoning using invisible stage tokens.

    arXiv
    2610.12644
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    Llama-3.1-8B-Instruct, Qwen2.5-7B-Instruct, Mistral-7B-Instruct-v0.3
    AbstractHidden tokens plus a stage controller make the backdoor progressively corrupt reasoning upon triggering, and the authors believe that looking only at the answer will miss it.

    Cancerous Backdoor Attack changes LLM backdoors from one-time activation to multi-stage reasoning corruption that accumulates progressively with triggers.

    Paper analysisRead analysis
  2. AttacksarXiv 2610.09027

    P-VMI：对抗图像经 KV cache 持续影响多轮对话

    提出持久视觉记忆注入，使对抗图像经缓存在触发时操控输出

    arXiv
    2610.09027
    Published
    Layer
    Infrastructure layer
    Setting
    Models and APIs
    Evaluated models
    Qwen3-8B, Qwen3-VL-8B-Instruct, Gemma-4-12B

    Abstract论文提出 P-VMI，证实对抗输入被掩蔽后仍可经 KV 缓存持久控制模型，强调了保留状态的安全风险。

    Paper analysisRead analysis
  3. AttacksarXiv 2610.07723

    Researchers propose answer-side backdoors: letting the model self-generate trigger words to bypass safety refusals.

    提出回答端后门，让模型自生成触发词以在干净有害请求下绕过拒答

    arXiv
    2610.07723
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    Llama-2-7B-chat-hf, Mistral-7B-Instruct-v0.3, Gemma-7B-it and others4

    Abstract论文提出了多轮对话中回答端自生成触发词绕过安全拒绝的后门攻击，指出输入端防护存在盲区。

    Paper analysisRead analysis
  4. AttacksarXiv 2609.33985

    研究：众包微调数据投毒可放大专有指令抽取

    提出主题锚点投毒，放大众包微调后的专有指令提取

    arXiv
    2609.33985
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    Qwen2.5-7B-Instruct, Qwen2.5-14B-Instruct, Llama-3.1-8B-Instruct and others4

    Abstract论文提出基于主题锚点的数据中毒攻击，证明低比例的中毒样本可在黑盒条件下隐蔽放大微调指令泄漏。

    Paper analysisRead analysis
  6. AttacksarXiv 2608.09577

    ElasticBack：通过触发词与规则耦合在 LLM Agent 技能中植入条件后门

    提出 ElasticBack，在单个 Agent 技能中植入条件后门

    arXiv
    2608.09577
    Published
    Layer
    Application layer
    Setting
    coding agent
    Evaluated models
    GLM-5.2, MiniMax-M3, GPT-5.4 and others4
    AbstractElasticBack 用冻结规则加遗传搜索演化 trigger，在单 skill 上做条件后门。

    ElasticBack 是放在 LLM agent 单个 skill 里的条件后门攻击。。

    Paper analysisRead analysis
You have reached the end