INDEX
    Explanations
    New Auto-Interp
    Negative Logits
    !/
    -0.09
     Audrey
    -0.09
     ypos
    -0.08
     Chinatown
    -0.08
     obt
    -0.07
     she
    -0.07
     persecution
    -0.07
    arr
    -0.07
     elde
    -0.07
     arro
    -0.07
    POSITIVE LOGITS
     knobs
    0.08
     المعرفة
    0.08
     gates
    0.08
     Babel
    0.08
     modifications
    0.08
     personalization
    0.08
     modifying
    0.08
    ritable
    0.08
     글로벌
    0.08
    <J
    0.08
    Act Density 0.002%

    No Known Activations