INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     arac
    -0.08
     Tanner
    -0.08
     sul
    -0.08
    .Reference
    -0.08
    .Power
    -0.08
     woods
    -0.08
     Finish
    -0.08
    -0.07
     wood
    -0.07
    -0.07
    POSITIVE LOGITS
    rather
    0.08
    MU
    0.08
    Assim
    0.08
    /filter
    0.08
     gestaltet
    0.08
     niż
    0.08
     daripada
    0.08
    andal
    0.08
    mu
    0.07
    mín
    0.07
    Act Density 0.022%

    No Known Activations