INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     frantic
    -0.08
     зы
    -0.08
    %">
    -0.08
    //'
    -0.08
     Tud
    -0.08
    \t
    -0.08
     EXIST
    -0.07
     draai
    -0.07
     violence
    -0.07
     velvet
    -0.07
    POSITIVE LOGITS
     gradients
    0.10
     progression
    0.08
     socks
    0.08
     progressed
    0.08
    carbonate
    0.08
    grat
    0.08
    Progress
    0.07
     analges
    0.07
    تاح
    0.07
     الحصول
    0.07
    Act Density 0.002%

    No Known Activations