Skip to content

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

Oct 12 · Monday

All papers

Found 1 paper
Filter4

Attack type

Jailbreaks1Poisoning and backdoors17Model tampering10Extraction and theft5

Attack surface / entry point

User input17Documents and messages0Weights and gradients1

Layer

Model layer5Training and data layer1
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 254papers have no tags yet and are not included in filtered results.
Attack type：JailbreaksAttack surface / entry point：Documents and messagesAttack surface / entry point：Weights and gradientsLayer：Training and data layerClear all

Another 254papers do not have category tags yet and are not included in filtered results.

  1. DefensesarXiv 2609.02293

    SEAL：通过共享专家对齐强化 MoE 全局安全

    提出 SEAL，在 MoE 共享专家上做 DPO-LoRA 对齐以抵御越狱与权重篡改

    arXiv
    2609.02293
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Attacker
    White-box, Black-box
    Evaluated models
    Qwen1.5-MoE-A2.7B, DeepSeekMoE-16B, GLM-4.7-Flash and others4
    AbstractSEAL 把 DPO-LoRA 限制在混合 MoE 共享专家上，SEAL++再保护已有安全神经元方向。

    SEAL是面向混合 MoE 的训练时防御：只对齐始终执行的共享专家，使安全不依赖路由器是否把 token 送到安全专家。

    Paper analysisRead analysis
You have reached the end