INDEX
    Explanations
    New Auto-Interp
    Negative Logits
    適切な
    0.42
     Effectively
    0.41
    ánea
    0.39
     lightest
    0.39
     neler
    0.37
     эффективно
    0.37
     efet
    0.36
    effective
    0.36
    Effective
    0.36
     дальней
    0.36
    POSITIVE LOGITS
     immensely
    1.72
     tremendously
    1.45
     dearly
    1.44
     enormously
    1.42
     greatly
    1.26
     deeply
    1.16
     hugely
    1.10
     profundamente
    1.08
     intensely
    1.02
     terribly
    0.98
    Act Density 0.046%

    No Known Activations