INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     cof
    -0.09
    Fab
    -0.09
    Eso
    -0.08
    NZ
    -0.08
    Hat
    -0.08
    Pf
    -0.08
     centros
    -0.07
    oolean
    -0.07
     recorrer
    -0.07
    Ft
    -0.07
    POSITIVE LOGITS
    0.09
     inhal
    0.08
     aconte
    0.08
     addiction
    0.08
     baños
    0.08
     dearly
    0.08
     атмосфер
    0.08
    失败
    0.08
    进去
    0.08
    0.08
    Act Density 0.001%

    No Known Activations