INDEX
    Explanations

    edges, surfaces, outlines

    New Auto-Interp
    Negative Logits
     behaviors
    0.38
     behaviours
    0.38
     ডাকা
    0.37
     удалось
    0.35
    0.35
     озе
    0.35
     माफ
    0.34
    outines
    0.34
     Behav
    0.34
     واقعات
    0.34
    POSITIVE LOGITS
     edges
    0.90
     sides
    0.73
     contours
    0.63
     underside
    0.58
     outline
    0.58
     tops
    0.57
     edge
    0.57
     corners
    0.57
     tips
    0.56
     insides
    0.56
    Act Density 0.083%

    No Known Activations