INDEX
    Explanations

    sizes, proportions, quantities

    New Auto-Interp
    Negative Logits
    oretically
    0.53
    odore
    0.52
    ophylline
    0.49
     gist
    0.46
     easiest
    0.45
     following
    0.45
     culprit
    0.44
    lma
    0.43
     simplest
    0.42
     afore
    0.41
    POSITIVE LOGITS
     размеров
    0.40
    struktur
    0.39
     விகி
    0.38
     Algunos
    0.37
    サイズ
    0.37
    边界
    0.37
     beberapa
    0.36
    0.36
     quelques
    0.36
     algumas
    0.36
    Act Density 0.004%

    No Known Activations