INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     Thanks
    0.54
     THANK
    0.54
     gratitude
    0.53
    感謝
    0.52
     Thankfully
    0.50
    感谢
    0.50
     Thank
    0.50
     спасибо
    0.49
     thankful
    0.49
     благодар
    0.49
    POSITIVE LOGITS
     but
    0.74
     nhưng
    0.69
    Sorry
    0.63
     mutta
    0.63
     ngunit
    0.63
     pero
    0.60
    Unable
    0.59
    but
    0.58
     लेकिन
    0.58
     unable
    0.57
    Act Density 0.004%

    No Known Activations