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

    Neuronpedia

    INDEX
    Explanations

    asking what

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    া
    1.15
     постро
    1.09
     возмо
    1.05
     добав
    1.04
    ки
    1.02
     до
    1.02
     ήταν
    1.01
     необходи
    0.99
    カイブ
    0.99
    াতি
    0.97
    POSITIVE LOGITS
    কে
    1.52
    '
    1.13
    \
    1.13
     what
    1.07
    t
    1.02
    ，
    0.96
     as
    0.95
     Unternehmen
    0.95
    。\
    0.93
    }=\
    0.93
    Act Density 0.208%

    No Known Activations