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

    Neuronpedia

    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Negative Logits
     lär
    -1.00
     länge
    -0.92
    𝙽
    -0.91
     swiftly
    -0.90
     två
    -0.89
    우스
    -0.89
     trän
    -0.88
    INOS
    -0.87
     vienas
    -0.86
    امت
    -0.85
    POSITIVE LOGITS
     comprises
    1.09
    たくない
    1.05
     –
    1.05
     $-$
    1.01
     granat
    0.93
     στην
    0.89
    ligan
    0.89
     bombard
    0.88
     stats
    0.86
     alterações
    0.86
    Act Density 0.000%

    No Known Activations

    This feature has no known activations.