INDEX
    Explanations
    New Auto-Interp
    Negative Logits
    さまざまな
    1.25
     delightfully
    1.09
     convivial
    1.04
     decidedly
    1.00
     deftly
    0.99
     variously
    0.97
     чрезвы
    0.95
     fortuit
    0.92
     succinctly
    0.91
     нередко
    0.90
    POSITIVE LOGITS
     ppl
    2.14
     yg
    1.96
     govt
    1.76
     fucking
    1.67
     shitty
    1.65
     tbh
    1.64
     đc
    1.63
     idk
    1.61
     ceo
    1.61
     gpu
    1.60
    Act Density 0.615%

    No Known Activations