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

    Neuronpedia

    INDEX
    Explanations

    when, while, this

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    <unused919>
    1.11
     tamaño
    1.04
    <unused932>
    0.99
    <unused719>
    0.99
    <unused682>
    0.99
    <unused1114>
    0.98
    <unused965>
    0.97
    <unused1644>
    0.97
    <unused187>
    0.97
    <unused670>
    0.97
    POSITIVE LOGITS
     ise
    0.73
     while
    0.72
     when
    0.69
     When
    0.68
    T
    0.66
    后者
    0.66
    ;
    0.65
     نیز
    0.65
    +,
    0.65
    when
    0.64
    Act Density 0.260%

    No Known Activations