INDEX
    Explanations

    Japanese honorifics/media

    New Auto-Interp
    Negative Logits
    yg
    -0.07
    dx
    -0.07
    	expected
    -0.07
     repr
    -0.07
    orp
    -0.07
    interest
    -0.07
     Sequelize
    -0.07
     prop
    -0.07
     challenging
    -0.07
     Dy
    -0.07
    POSITIVE LOGITS
    šie
    0.08
    哥哥
    0.08
     oudere
    0.08
    пі
    0.08
    0.08
    ৃত
    0.08
    0.08
     aʻe
    0.08
    әк
    0.08
    ’,
    0.07
    Act Density 0.007%

    No Known Activations