INDEX
    Explanations

    too / enough

    New Auto-Interp
    Negative Logits
     Transform
    -0.08
     gle
    -0.08
     ph
    -0.08
    iente
    -0.07
     Ph
    -0.07
    েক্ষ
    -0.07
     handed
    -0.07
     Sprach
    -0.07
    -transform
    -0.07
    Prest
    -0.07
    POSITIVE LOGITS
     Bethlehem
    0.08
     ګڼ
    0.08
    及时
    0.08
     احساس
    0.08
    rels
    0.08
     లో
    0.08
    房地产
    0.08
     musul
    0.08
    0.08
     хватает
    0.08
    Act Density 0.035%

    No Known Activations