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

    Neuronpedia

    INDEX
    Explanations

    Crane

    np_max-act · gemini-2.0-flash

    tokens that mark assistant-produced text or the assistant speaker role in the conversation.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    弯曲
    -0.07
     totalement
    -0.07
     העל
    -0.07
     NE
    -0.06
    狀
    -0.06
     Niet
    -0.06
    🏳
    -0.06
    微量
    -0.06
    tatus
    -0.06
    _backward
    -0.06
    POSITIVE LOGITS
    �
    0.08
    .ComboBox
    0.07
    淮
    0.06
    沈阳
    0.06
     CARD
    0.06
    ancing
    0.06
    .Par
    0.06
     Văn
    0.06
     Francesco
    0.06
    /tests
    0.06
    Act Density 0.013%

    No Known Activations