INDEX
    Explanations

    The neuron fires on numeric identifier tokens—especially version numbers or article IDs embedded in MSDN/documentation URLs.

    New Auto-Interp
    Negative Logits
    upertino
    -0.06
     Steve
    -0.06
    法国
    -0.06
     Karen
    -0.06
     Fire
    -0.06
     كتب
    -0.06
     HH
    -0.06
    -0.06
    /mail
    -0.06
    _images
    -0.06
    POSITIVE LOGITS
     '',
    0.07
    ieved
    0.07
     dif
    0.07
    );}↵
    0.07
    	uint
    0.06
    illum
    0.06
    бут
    0.06
    _DIM
    0.06
    _glob
    0.06
    GRAPH
    0.06
    Act Density 0.003%

    No Known Activations