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

    Neuronpedia

    INDEX
    Explanations

    2

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     precious
    1.31
     valuable
    1.18
     delicate
    1.14
     sensitive
    1.06
     fragile
    0.98
     expensive
    0.97
     vulnerable
    0.96
     invaluable
    0.91
     finite
    0.90
     pricey
    0.90
    POSITIVE LOGITS
    e
    0.85
     तेव्हा
    0.80
     <!--
    0.76
    isering
    0.76
    isasi
    0.76
    *',
    0.74
     тў
    0.74
     větš
    0.74
    체를
    0.71
    체
    0.71
    Act Density 0.173%

    No Known Activations