INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     presumably
    0.47
    امات
    0.43
     Presumably
    0.40
     medieval
    0.39
     anlamına
    0.39
     የበ
    0.38
     будто
    0.38
     Bratislava
    0.37
    Sap
    0.37
     boast
    0.36
    POSITIVE LOGITS
     think
    0.80
     thinks
    0.75
    think
    0.67
     suspect
    0.67
     firmly
    0.67
     believe
    0.66
     Think
    0.63
     personally
    0.63
     nghĩ
    0.62
     believes
    0.61
    Act Density 0.004%

    No Known Activations