INDEX
    Explanations

    references to various ethnic and religious groups

    New Auto-Interp
    Negative Logits
    jee
    -0.17
    gger
    -0.16
    .gg
    -0.15
    fad
    -0.15
     Downing
    -0.15
    639
    -0.14
    人åijĺ
    -0.14
    dings
    -0.14
    gest
    -0.14
    iglia
    -0.14
    POSITIVE LOGITS
     who
    0.27
    who
    0.20
     whom
    0.20
    -American
    0.20
    -Americans
    0.17
     اÙĦذÙĬÙĨ
    0.17
     kteÅĻÃŃ
    0.16
    /OR
    0.15
    /left
    0.15
    innen
    0.15
    Act Density 0.091%

    No Known Activations