INDEX
    Explanations

    hallucinations

    New Auto-Interp
    Negative Logits
     جنوبی
    -0.07
     ukon
    -0.07
    /index
    -0.07
     Borders
    -0.06
     Cedar
    -0.06
     Cres
    -0.06
    -med
    -0.06
    BTTagCompound
    -0.06
     reproductive
    -0.06
    ulist
    -0.06
    POSITIVE LOGITS
     halluc
    0.09
     recalled
    0.07
    Did
    0.06
    SourceType
    0.06
     blob
    0.06
     Images
    0.06
     студ
    0.06
     tortured
    0.06
    controlled
    0.06
    danger
    0.06
    Act Density 0.005%

    No Known Activations