INDEX
    Explanations

    python code

    New Auto-Interp
    Negative Logits
     cockpit
    -0.08
     prestige
    -0.08
    apas
    -0.08
     Worse
    -0.08
    Founder
    -0.07
     exclusivement
    -0.07
     victime
    -0.07
     robes
    -0.07
    Pago
    -0.07
     sophistication
    -0.07
    POSITIVE LOGITS
     yields
    0.11
     yield
    0.09
    0.09
     yielding
    0.08
     queries
    0.08
     ergibt
    0.08
    ::↵↵
    0.08
     homo
    0.08
     yielded
    0.08
     ش
    0.08
    Act Density 0.017%

    No Known Activations