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

    Neuronpedia

    INDEX
    Explanations

    opens up possibilities

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ﺑ
    0.86
    ění
    0.83
    cout
    0.81
    相比
    0.77
    buf
    0.76
    %\
    0.75
     usu
    0.75
     органов
    0.74
    ඹ
    0.74
    strom
    0.73
    POSITIVE LOGITS
     doors
    1.66
     up
    1.55
     Doors
    1.30
    Doors
    1.27
    doors
    1.25
     sores
    1.18
     doorways
    1.17
     sesame
    1.16
     gamb
    1.16
     gates
    1.13
    Act Density 0.041%

    No Known Activations