INDEX
    Explanations

    structural elements in formatted documents, such as figures and captions

    New Auto-Interp
    Negative Logits
    uels
    -0.16
    loi
    -0.15
    aits
    -0.15
    оÑĢаз
    -0.15
    datal
    -0.15
    lost
    -0.14
    ],[-
    -0.14
     Reform
    -0.14
    lech
    -0.13
    andles
    -0.13
    POSITIVE LOGITS
    TypeInfo
    0.16
    Äįky
    0.15
    Äĥ
    0.15
    ObjectName
    0.15
    chia
    0.14
    rig
    0.14
    ONTAL
    0.14
    ιο
    0.14
     Flush
    0.14
    541
    0.14
    Act Density 0.017%

    No Known Activations