Skip to content

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

Oct 12 · Monday

All papers

Found 4 papers
Filter5

Attack type

Poisoning and backdoors3

Impact

Unauthorized actions4

System type

Models and APIs0AI Agent58coding agent4web agent15Browser and computer use5Multi-agent systems4

Attack surface / entry point

User input10Web pages and interfaces1Tool responses4Tool and skill descriptions16Code repositories and dependencies10Memory writes3Distribution channels11

Layer

Application layer4
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.
Impact：Unauthorized actionsSystem type：Models and APIsSystem type：coding agentAttack surface / entry point：Web pages and interfacesAttack surface / entry point：Memory writesClear all

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

  2. Evaluation and benchmarksarXiv 2609.32616

    Cave-Bench：虚假指控下 LLM 智能体最高 60.06% 的运行会破坏正确工作

    提出 CAVE-BENCH，评测智能体在虚假指责下破坏已完成的正确工作

    arXiv
    2609.32616
    Published
    Layer
    Application layer
    Setting
    coding agent
    Evaluated models
    Claude-Sonnet-5, Claude-Opus-5, GPT-5.6-Sol and others14

    Abstract论文提出了评估智能体在虚假指责下破坏已有正确成果的 CAVE-BENCH，发现强模型易推翻已知证据，提出了门控缓解方案。

    Paper analysisRead analysis
You have reached the end