INDEX
    Explanations

    specific special characters or diacritics in text

    New Auto-Interp
    Negative Logits
    bane
    -0.17
    ago
    -0.16
    دÙĩÙħ
    -0.15
    anges
    -0.15
    ré
    -0.14
    enames
    -0.14
    ugi
    -0.14
    inspect
    -0.14
    eness
    -0.14
    atur
    -0.14
    POSITIVE LOGITS
    cher
    0.16
    اÙĨÙĩ
    0.15
     Pert
    0.15
    ffen
    0.15
    ieux
    0.15
    ött
    0.14
    glm
    0.14
    edy
    0.14
    lichen
    0.14
    ÄŁÃ¼
    0.14
    Act Density 0.024%

    No Known Activations