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

    Neuronpedia

    INDEX
    Explanations

    end of shell conditional blocks

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     and
    0.86
    ered
    0.72
    END
    0.70
    ordine
    0.68
     which
    0.67
    ched
    0.66
    IFY
    0.66
    ﺷ
    0.66
    TING
    0.66
    //
    0.65
    POSITIVE LOGITS
    ب
    1.17
    ی
    0.94
    ॅमिली
    0.92
    ী
    0.90
    ف
    0.88
    k
    0.87
    i
    0.86
    ه
    0.83
    t
    0.82
    ج
    0.81
    Act Density 1.606%

    No Known Activations