INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     אש
    -0.08
    _pass
    -0.07
     Seeing
    -0.07
     Stil
    -0.07
     Trips
    -0.07
    ştır
    -0.07
     ব্যব
    -0.07
    _insert
    -0.07
    örn
    -0.07
    ированный
    -0.07
    POSITIVE LOGITS
    0.20
     contour
    0.14
    0.12
     contours
    0.12
    (cont
    0.12
     contou
    0.11
    Contour
    0.11
     outline
    0.10
    0.10
    cont
    0.10
    Act Density 0.002%

    No Known Activations