Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    AI

    np_max-act · gemini-2.0-flash

    Instances where the assistant self-identifies or gives a disclaimer about being an AI (the model's "As an AI ..." style preface).

    oai_token-act-pair · gpt-5-miniTriggered by @yooniel31
    New Auto-Interp
    Negative Logits
    ystone
    -0.07
     ung
    -0.06
    otre
    -0.06
     مق
    -0.06
    距
    -0.06
    vals
    -0.06
     Γεω
    -0.06
     repairs
    -0.06
     bilg
    -0.06
    飞
    -0.06
    POSITIVE LOGITS
    (cv
    0.06
    Eth
    0.06
    \Service
    0.06
    _references
    0.06
     Eth
    0.06
    *(-
    0.06
    �
    0.06
    本
    0.06
    ／:
    0.06
    ธ
    0.06
    Act Density 0.062%

    No Known Activations