INDEX
    Explanations

    references to content located "above" a certain point in the document

    New Auto-Interp
    Negative Logits
     Schul
    -0.16
    ers
    -0.16
    ling
    -0.15
    /display
    -0.15
    wid
    -0.15
     widely
    -0.14
     Trib
    -0.14
    odega
    -0.14
    ob
    -0.14
     Sou
    -0.14
    POSITIVE LOGITS
    .scalablytyped
    0.18
    abay
    0.17
    atoria
    0.16
    /Dk
    0.16
    ofire
    0.15
    datable
    0.15
     YYS
    0.15
    zes
    0.14
    alah
    0.14
     î¡
    0.14
    Act Density 0.029%

    No Known Activations