INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     asleep
    -0.08
    motion
    -0.08
     uusi
    -0.08
     sul
    -0.07
     unreliable
    -0.07
    τας
    -0.07
    orrow
    -0.07
     madd
    -0.07
     family's
    -0.07
    ös
    -0.07
    POSITIVE LOGITS
     trabalham
    0.08
    hetically
    0.07
    .pack
    0.07
    jetër
    0.07
    Side
    0.07
     FITNESS
    0.07
    hetics
    0.07
     leder
    0.07
    вер
    0.07
     Zwe
    0.07
    Act Density 0.001%

    No Known Activations