INDEX
    Explanations

    words related to metadata such as names, dates, and locations

    the presence of vertical bar characters (|) indicating section or segment delimiters

    New Auto-Interp
    Negative Logits
    anski
    -0.86
    ifts
    -0.83
    rons
    -0.82
    ory
    -0.80
    ippi
    -0.80
    anium
    -0.80
    alis
    -0.80
    eele
    -0.76
    orical
    -0.75
    raints
    -0.74
    POSITIVE LOGITS
     |--
    0.96
    cffff
    0.95
    ··
    0.91
    âĢ¢âĢ¢
    0.73
     Reached
    0.72
    ————
    0.71
    âĢ¢âĢ¢âĢ¢âĢ¢
    0.71
    cffffcc
    0.71
     EntityItem
    0.70
    phabet
    0.69
    Act Density 0.015%

    No Known Activations