INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     olho
    1.35
     opsi
    1.30
     pilih
    1.27
     olduğu
    1.27
     sljede
    1.23
     keempat
    1.19
     avendo
    1.17
     olursa
    1.17
     caranya
    1.16
     escolher
    1.16
    POSITIVE LOGITS
    <eos>
    1.71
    1.17
    ↵↵
    1.08
    eter
    0.95
    </h1>
    0.95
    ↵↵↵↵
    0.94
    ↵↵↵
    0.91
    ↵↵↵↵↵↵
    0.90
    <0x0C>
    0.89
    </blockquote>
    0.89
    Act Density 0.242%

    No Known Activations