INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     epsilon
    -0.10
     eps
    -0.09
     election
    -0.08
    -0.07
     Par
    -0.07
     dividend
    -0.07
     advantage
    -0.07
     ε
    -0.07
     микро
    -0.07
    .Flush
    -0.07
    POSITIVE LOGITS
     undes
    0.10
     unintended
    0.09
    undes
    0.09
     undesirable
    0.08
    iefs
    0.08
    0.08
     glor
    0.08
     fascination
    0.08
     проник
    0.08
     Anatom
    0.08
    Act Density 0.029%

    No Known Activations