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

    Neuronpedia

    INDEX
    Explanations

    understanding reasons why

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     how
    2.63
     why
    2.25
    how
    1.90
    why
    1.87
     what
    1.87
     whether
    1.87
    为什么
    1.80
     How
    1.75
    如何
    1.68
    どのように
    1.65
    POSITIVE LOGITS
     مطابق
    0.85
     saya
    0.84
     ሶ
    0.83
     segera
    0.82
     両
    0.79
     eléctrica
    0.78
     sekali
    0.76
    ocèle
    0.76
    csvStream
    0.76
    acağız
    0.75
    Act Density 0.110%

    No Known Activations