INDEX
    Explanations
    New Auto-Interp
    Negative Logits
    irectional
    -0.08
    गे
    -0.08
    word
    -0.08
     companionship
    -0.08
     spiel
    -0.08
    Pocket
    -0.07
    Gew
    -0.07
     rode
    -0.07
     shortlist
    -0.07
    (run
    -0.07
    POSITIVE LOGITS
     частью
    0.09
     ven
    0.08
    /is
    0.08
     stata
    0.08
     ضر
    0.08
     CW
    0.08
     og
    0.07
     отзыв
    0.07
    0.07
     представ
    0.07
    Act Density 0.032%

    No Known Activations