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

    Neuronpedia

    INDEX
    Explanations

    Inhaling, Indonesian, Chinese, Russian

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ০
    9.75
    in
    9.14
    a
    8.92
    ي
    8.56
    ת
    8.50
    i
    8.50
    an
    7.80
    u
    7.76
    er
    7.66
    y
    7.62
    POSITIVE LOGITS
    了一个
    3.48
    bbene
    2.93
    Кто
    2.80
    Когда
    2.73
    了一
    2.68
    Най
    2.62
    𝐭
    2.61
    Итак
    2.59
    了一個
    2.57
    ibouti
    2.56
    Act Density 0.850%

    No Known Activations