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

    Neuronpedia

    INDEX
    Explanations

    class member names

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    bsen
    -1.11
    masing
    -1.03
    unno
    -0.94
    endforeach
    -0.94
    nad
    -0.92
    lios
    -0.91
    tiamo
    -0.91
    USAL
    -0.91
    vereignty
    -0.88
     tos
    -0.88
    POSITIVE LOGITS
     would
    1.06
     might
    1.01
     because
    0.96
    ándome
    0.89
    larının
    0.84
     what
    0.83
     actually
    0.83
    蔭
    0.80
     other
    0.79
    డ్
    0.79
    Act Density 0.009%

    No Known Activations