INDEX
    Explanations

    phrases containing symbols (e.g., âĢ, ľ) commonly used for emphasis or decoration

    the presence of end-of-text tokens indicating the completion of sections or ideas

    New Auto-Interp
    Negative Logits
     gad
    -0.72
     scattering
    -0.71
     dispers
    -0.70
    anwhile
    -0.69
    ierrez
    -0.68
     casting
    -0.68
     scatter
    -0.66
     nearest
    -0.66
     detached
    -0.65
     Peb
    -0.64
    POSITIVE LOGITS
    º
    1.23
    ¹
    1.11
    £
    1.08
    ®
    1.05
    į
    1.04
    ı
    1.02
    Į
    1.02
    Ī
    1.00
    ¦
    0.98
    ¬
    0.96
    Act Density 0.107%

    No Known Activations