INDEX
    Explanations

    Numerical word problems

    New Auto-Interp
    Negative Logits
    оби
    -0.07
     biting
    -0.07
     konk
    -0.07
     behaupt
    -0.07
    awi
    -0.07
     deten
    -0.07
    Liv
    -0.07
     тод
    -0.07
    gear
    -0.07
    σκ
    -0.07
    POSITIVE LOGITS
     interplay
    0.09
     issus
    0.08
    办理
    0.07
    .cod
    0.07
     provenant
    0.07
    0.07
    (accounts
    0.07
     doubl
    0.07
     Nin
    0.07
    (Transaction
    0.07
    Act Density 0.098%

    No Known Activations