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

    Neuronpedia

    INDEX
    Explanations

    cracks knuckles, cracked plaster

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    <0x80>
    1.13
    ناک
    1.06
    یم
    1.02
    ку
    0.95
    ڈر
    0.93
    یسم
    0.93
    ้
    0.93
    বা
    0.92
    کی
    0.92
    ża
    0.92
    POSITIVE LOGITS
    et
    1.29
    st
    1.20
    s
    1.17
    О
    1.13
    2
    1.06
    e
    1.04
    _
    1.03
    ه
    0.98
     Crack
    0.98
    1
    0.98
    Act Density 0.006%

    No Known Activations