INDEX
    Explanations
    New Auto-Interp
    Negative Logits
    beam
    -0.08
     Libya
    -0.08
     économies
    -0.08
     beam
    -0.08
    ipay
    -0.08
     Basin
    -0.08
    osse
    -0.08
     nám
    -0.08
    onen
    -0.08
    avoir
    -0.08
    POSITIVE LOGITS
    Pseudo
    0.08
     pseudo
    0.07
     Fake
    0.07
     Supervis
    0.07
     quoting
    0.07
     allegedly
    0.07
     alleged
    0.07
    QUOTE
    0.07
    .struct
    0.07
    WARD
    0.07
    Act Density 0.081%

    No Known Activations