INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     बैठ
    -0.08
     Cur
    -0.08
     curvature
    -0.08
    .Alter
    -0.08
    -0.07
    -0.07
     विशेष
    -0.07
     funcionamento
    -0.07
     contamin
    -0.07
     accord
    -0.07
    POSITIVE LOGITS
     baka
    0.08
     crazy
    0.08
     отказ
    0.08
     fuck
    0.08
    Lean
    0.07
    iken
    0.07
    яд
    0.07
    ìn
    0.07
     towards
    0.07
     clam
    0.07
    Act Density 0.013%

    No Known Activations