INDEX
    Explanations

    words before specific topics

    New Auto-Interp
    Negative Logits
    确实
    0.54
    行って
    0.47
    भाष
    0.47
    Granted
    0.45
     मिळा
    0.45
     قوت
    0.43
     prid
    0.43
     partiellement
    0.43
    Indeed
    0.42
    0.42
    POSITIVE LOGITS
     hypertext
    0.79
     Socialism
    0.73
     airfoil
    0.70
     orthodont
    0.70
     saxophone
    0.69
     cyberpunk
    0.69
     ballroom
    0.68
     tensors
    0.67
     socialism
    0.67
     tattoos
    0.67
    Act Density 0.430%

    No Known Activations