INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     Brook
    -0.07
     ما
    -0.07
     observers
    -0.07
    Doctrine
    -0.07
    'ob
    -0.07
     Abby
    -0.07
    Bor
    -0.07
     Далее
    -0.07
     Glock
    -0.07
     Влад
    -0.07
    POSITIVE LOGITS
    .encrypt
    0.07
    しょう
    0.07
    .layer
    0.07
     @(
    0.07
    0.07
    arena
    0.07
     affine
    0.07
     juice
    0.07
    0.07
     exponentially
    0.07
    Act Density 0.001%

    No Known Activations