INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     anyway
    0.65
     guesswork
    0.63
     vines
    0.63
     plus
    0.63
     blocking
    0.62
     if
    0.61
     via
    0.61
    Scale
    0.61
     harmonics
    0.60
     scale
    0.60
    POSITIVE LOGITS
     अत्यंत
    1.13
     Öncelikle
    1.10
     অত্যন্ত
    1.10
     extrêmement
    1.08
     extremamente
    1.02
     ખૂબ
    1.00
     очень
    0.99
     estremamente
    0.97
     дуже
    0.96
     ძალიან
    0.96
    Act Density 0.466%

    No Known Activations