INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     fortune
    -0.08
    Lon
    -0.07
     Lum
    -0.07
     surged
    -0.07
    ancy
    -0.07
    _lon
    -0.07
    fortune
    -0.07
    quoi
    -0.07
    ët
    -0.07
     Astronomy
    -0.07
    POSITIVE LOGITS
    0.08
    -knit
    0.07
    0.07
    0.07
     mattered
    0.07
    wach
    0.07
     সবাই
    0.07
    cracker
    0.07
    гой
    0.07
    וג
    0.07
    Act Density 0.002%

    No Known Activations