INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     Marm
    -0.08
     malfunction
    -0.08
     marm
    -0.08
    -0.07
    .operation
    -0.07
    -0.07
    @\
    -0.07
    've
    -0.07
    @"
    -0.07
    struction
    -0.07
    POSITIVE LOGITS
     fo
    0.09
    anner
    0.08
     Pars
    0.08
    -looking
    0.08
    -valu
    0.08
    ‌تر
    0.08
     الحجم
    0.07
    -ish
    0.07
    0.07
    itele
    0.07
    Act Density 0.005%

    No Known Activations