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

    Neuronpedia

    INDEX
    Explanations

    transition words followed by commas

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     nevertheless
    1.21
     portanto
    1.08
     പക്ഷേ
    1.06
     kuitenkin
    1.04
     nonetheless
    1.03
     tehát
    1.03
     porém
    1.02
     बेशक
    1.00
     totiž
    1.00
     ovviamente
    1.00
    POSITIVE LOGITS
    ك
    1.01
    ﻙ
    0.95
    ﺏ
    0.92
     Есть
    0.86
     recent
    0.82
    様な
    0.81
    ुप
    0.80
    ్వ
    0.80
    ﻥ
    0.79
    ک
    0.79
    Act Density 0.073%

    No Known Activations