INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     perpetrated
    0.59
     disillusioned
    0.58
     unjustly
    0.58
     финансо
    0.57
     informée
    0.57
     comorbid
    0.57
     politiques
    0.56
     altru
    0.56
    🧠
    0.56
     activism
    0.55
    POSITIVE LOGITS
     straps
    0.89
     plastic
    0.89
     sleeves
    0.88
     flanges
    0.88
     wooden
    0.87
     cylindrical
    0.84
     ribbed
    0.84
     strainer
    0.84
     washers
    0.84
     removable
    0.83
    Act Density 0.101%

    No Known Activations