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

    Neuronpedia

    INDEX
    Explanations

    certified, quantization

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ী
    3.99
    го
    3.02
    ية
    2.99
    м
    2.95
    ش
    2.92
    spaper
    2.90
    ant
    2.87
    ie
    2.66
    д
    2.61
    특별시
    2.57
    POSITIVE LOGITS
    eh
    3.31
    ei
    3.05
    ej
    2.80
    Архівовано
    2.74
    $=
    2.69
    /"+
    2.60
    ekkür
    2.60
    ۰۰
    2.59
    $=\
    2.57
    eine
    2.56
    Act Density 0.029%

    No Known Activations