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

    Neuronpedia

    INDEX
    Explanations

    names followed by surnames

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     real
    0.63
    eval
    0.61
     quite
    0.59
     more
    0.59
     typically
    0.58
     ideally
    0.58
      
    0.57
    com
    0.56
     percentages
    0.56
     things
    0.56
    POSITIVE LOGITS
     J
    1.04
    <unused1207>
    0.97
     citado
    0.94
     डब्ल्यू
    0.91
    chyné
    0.91
    <unused1228>
    0.91
    <unused1087>
    0.90
    <unused1893>
    0.90
     señala
    0.90
    <unused379>
    0.90
    Act Density 0.045%

    No Known Activations