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

    Neuronpedia

    INDEX
    Explanations

    protecting or safeguarding

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    "
    1.10
    ı
    1.00
    ;
    0.85
     poderá
    0.82
     позволи
    0.82
    ”
    0.78
     aument
    0.78
     απε
    0.77
    '
    0.77
     is
    0.76
    POSITIVE LOGITS
    ל
    1.26
    ד
    1.23
    ن
    1.10
    י
    1.09
    н
    1.06
    д
    1.06
     for
    1.05
    ت
    1.04
    د
    1.02
    ק
    1.01
    Act Density 0.019%

    No Known Activations