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

    Neuronpedia

    INDEX
    Explanations

    normalization of numerical values

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     eighteen
    0.65
     Dragon
    0.65
     сх
    0.64
     Magic
    0.64
     ۱۵
    0.61
     eleven
    0.60
     seventeen
    0.59
     Indie
    0.59
     ab
    0.59
     Tor
    0.58
    POSITIVE LOGITS
     године
    1.00
    effetto
    0.92
    žno
    0.89
    Radians
    0.89
    ouwd
    0.89
    Giá
    0.87
     obesidad
    0.85
     godine
    0.85
    中华
    0.84
    <unused632>
    0.82
    Act Density 0.033%

    No Known Activations