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

    Neuronpedia

    INDEX
    Explanations

    behavior and its contexts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    م
    1.13
    ح
    1.11
    س
    1.09
    с
    1.04
    s
    0.96
    в
    0.96
    もら
    0.93
    ש
    0.92
    ی
    0.91
    ス
    0.91
    POSITIVE LOGITS
    ized
    1.00
    ки
    0.96
    ية
    0.93
    ığı
    0.92
    ,
    0.86
    стве
    0.85
    ac
    0.82
    할
    0.80
    вает
    0.79
    적인
    0.79
    Act Density 0.023%

    No Known Activations