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

    Neuronpedia

    INDEX
    Explanations

    defining where

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     God
    0.65
     بیشتری
    0.62
     Man
    0.60
    !
    0.60
     luck
    0.57
     fantastic
    0.57
     terrible
    0.57
    睹
    0.57
     fabulous
    0.56
    buzz
    0.56
    POSITIVE LOGITS
     where
    2.63
    where
    2.46
     Where
    2.30
     где
    2.28
     其中
    2.20
     donde
    2.19
    Where
    2.18
     où
    2.13
     حيث
    2.06
    其中
    2.03
    Act Density 0.168%

    No Known Activations