INDEX
    Explanations
    New Auto-Interp
    Negative Logits
    -0.07
     вел
    -0.07
    Epoch
    -0.07
     Bernard
    -0.07
     hei
    -0.07
     Henrik
    -0.07
     opět
    -0.07
    _restart
    -0.07
     ترتیب
    -0.07
     disappear
    -0.07
    POSITIVE LOGITS
     custom
    0.14
     Custom
    0.13
    Custom
    0.12
    custom
    0.12
    .Custom
    0.10
    /custom
    0.09
    -custom
    0.09
     cust
    0.08
     CUSTOM
    0.08
    (custom
    0.08
    Act Density 0.017%

    No Known Activations