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

    Neuronpedia

    INDEX
    Explanations

    utility classes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     trifling
    0.49
    Pages
    0.44
     hark
    0.43
     population
    0.42
    Number
    0.41
    population
    0.41
     denial
    0.40
    ุ
    0.40
     den
    0.40
     bit
    0.40
    POSITIVE LOGITS
    摶
    0.58
     duckys
    0.53
    奀
    0.50
     gà
    0.50
    tede
    0.50
    стаў
    0.50
    tempHeader
    0.49
    ستم
    0.49
    órica
    0.49
    iselt
    0.49
    Act Density 0.001%

    No Known Activations