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

    Neuronpedia

    INDEX
    Explanations

    capital, return, building

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    á
    0.71
     السياسي
    0.59
    يها
    0.59
     часу
    0.59
    PRF
    0.59
    FRA
    0.58
    ļu
    0.58
    的名
    0.58
    ம்
    0.57
     lång
    0.57
    POSITIVE LOGITS
     to
    0.89
     in
    0.82
     you
    0.82
    在
    0.81
     on
    0.80
     of
    0.73
    0
    0.73
    е
    0.72
    К
    0.70
     this
    0.68
    Act Density 0.000%

    No Known Activations