INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     iska
    -0.08
    leme
    -0.08
     dalla
    -0.08
     Hatch
    -0.08
     Dominican
    -0.07
     ata
    -0.07
    viel
    -0.07
     selecting
    -0.07
     Bj
    -0.07
     cyst
    -0.07
    POSITIVE LOGITS
     каж
    0.09
    க்
    0.08
     nitr
    0.08
     confid
    0.08
     Kansas
    0.07
     felis
    0.07
     Ni
    0.07
    allet
    0.07
    _perf
    0.07
    0.07
    Act Density 0.002%

    No Known Activations