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

    Neuronpedia

    INDEX
    Explanations

    following words like place names

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ம்
    2.36
    ка
    1.84
    س
    1.82
    ות
    1.73
    >
    1.73
     Dette
    1.68
    ють
    1.66
    ہ
    1.64
    ون
    1.63
    ס
    1.59
    POSITIVE LOGITS
     وفي
    2.30
     وبين
    2.05
    atrice
    1.87
    ا
    1.86
    ılarak
    1.81
    ı
    1.80
    ü
    1.70
    疚
    1.70
    부터
    1.69
    いきます
    1.68
    Act Density 0.656%

    No Known Activations