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

    Neuronpedia

    INDEX
    Explanations

    letter sequences at start of words

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ك
    1.99
    $
    1.45
    '
    1.37
    ING
    1.26
    س
    1.21
    ات
    1.17
    Foto
    1.17
    v
    1.16
    User
    1.14
    #
    1.13
    POSITIVE LOGITS
     Católica
    1.09
    롭게
    1.07
    ки
    0.98
    maßen
    0.98
    اً
    0.97
     adanya
    0.91
    を受けた
    0.86
     carácter
    0.85
    ෙහි
    0.85
     zugleich
    0.83
    Act Density 0.057%

    No Known Activations