INDEX
    Explanations

    design and technical tools

    This neuron detects capitalized technical acronyms and initialisms (especially security, medical imaging, AI, and enterprise software terms).

    New Auto-Interp
    Negative Logits
     
    0.63
     I
    0.52
    ۔
    0.46
    0.45
    ului
    0.40
    0.40
    0.39
    asia
    0.39
    kaan
    0.39
     decentral
    0.38
    POSITIVE LOGITS
    на
    1.02
     on
    0.74
    ו
    0.73
    و
    0.72
    to
    0.69
    ي
    0.67
    ح
    0.65
    ান
    0.63
    2
    0.62
    ف
    0.62
    Act Density 0.909%

    No Known Activations