INDEX
    Explanations

    bathroom features

    New Auto-Interp
    Negative Logits
    表达
    -0.10
     Plast
    -0.10
    -0.09
    importe
    -0.09
     Portuguesa
    -0.09
     hambre
    -0.09
    .RUNTIME
    -0.09
    .expr
    -0.08
     lazım
    -0.08
    -0.08
    POSITIVE LOGITS
     modeled
    0.08
     upgraded
    0.08
    -notch
    0.08
     splash
    0.07
     lounge
    0.07
     jpeg
    0.07
    oled
    0.07
     из
    0.07
     transitioned
    0.07
     lait
    0.07
    Act Density 0.012%

    No Known Activations