Skip to content

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

Oct 13 · Tuesday

All papers

Found 13 papers
Filter6

Attack technique

Gradient and representation optimization3

Model risks

Hallucination and fabrication3

Impact

Harmful content20Unauthorized actions8Data and privacy leaks9Output manipulation13Availability and reliability3Oversight failure24

System type

Models and APIs11LLM applications5AI Agent23coding agent2Multi-agent systems5

Attack surface / entry point

Images, audio and video7

Layer

Prompt layer4Model layer7

System component

Memory2Guardrails and judges13Monitors2Reasoning chains10Fine-tuning and open weights15Vision42Audio9Video6Image generation9Multimodal61
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 201papers have no tags yet and are not included in filtered results.
Impact：Output manipulationSystem type：Models and APIsSystem type：coding agentSystem component：MemorySystem component：MonitorsSystem component：AudioClear all

Another 201papers do not have category tags yet and are not included in filtered results.

  1. Evaluation and benchmarksarXiv 2610.06339

    BabelFake：面向多语言音视频 DeepFake 检测的基准数据集

    构建多语言音视频 DeepFake 基准 BabelFake 并评测检测器

    arXiv
    2610.06339
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    SpeechForensics, AVH-Align, (B)RAVEn
    AbstractBabelFake 为五语同意采集基准。

    BabelFake 是经同意与 IRB 批准采集的多语音视频 DeepFake 检测基准，用来把视觉操纵和语音合成拆开比较。

    Paper analysisRead analysis
  4. Evaluation and benchmarksarXiv 2609.39473

    FAME：用反事实推理评测自主智能体的虚假记忆

    提出 FAME，用反事实概念漂移评测智能体虚假记忆

    arXiv
    2609.39473
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Llama-3B
    AbstractFAME 用反事实引起的隐状态概念漂移，在出答案前区分忠实记忆与虚假记忆。

    FAME 是一个免训练的评测框架，用来判断自主智能体的记忆在虚假相关、环境偏移或知识冲突下是否已变成虚假记忆。。

    Paper analysisRead analysis
  5. DefensesarXiv 2609.37568

    SECRET：用源条件化中继引导缓解音视频大语言模型的幻觉

    提出训练无关的 SECRET，对问题表征做源条件转向以缓解音视频模型幻觉

    arXiv
    2609.37568
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    VideoLLaMA2-AV-7B, Qwen2.5-Omni-7B, Qwen3-Omni-30B-A3B
    AbstractSECRET 在问题状态上做源条件转向，三个 AVLLM 上减轻了跨模态接地幻觉。

    SECRET 是一种训练无关的问题状态转向方法，用来减轻 AVLLM 的 source-confused grounding hallucination。

    Paper analysisRead analysis
  8. AttacksarXiv 2609.01023

    Akrasia：针对推理型代码 LLM 的隐蔽后门攻击

    提出隐蔽后门攻击 AKRASIA，在推理代码模型提示中植入触发器并伪装思维链

    arXiv
    2609.01023
    Published
    Layer
    Prompt layer
    Setting
    coding agent
    Evaluated models
    Claude-Sonnet-5, GPT-5.5, Gemini-3.5-flash and others7

    Abstract论文针对推理代码大模型提出了隐蔽推断期后门攻击 AKRASIA，揭示了模型思维链可被利用进行欺骗性伪装的安全隐患。

    Paper analysisRead analysis
  12. AttacksarXiv 2604.14604

    The AudioHijack framework achieves covert audio prompt injection against 13 speech large models

    提出 AudioHijack，用不可感知对抗音频对语音模型做上下文无关提示注入

    arXiv
    2604.14604
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    SpeechGPT, GLM-4-Voice, VITA-Audio and others15

    Abstract提出了 AudioHijack 框架，揭示了 LALM 对听觉提示注入的脆弱性与传统防御的局限。

    Paper analysisRead analysis
You have reached the end