INDEX
    Explanations

    parentheses and numerical formats

    New Auto-Interp
    Negative Logits
     Verb
    -0.15
    lien
    -0.15
    &&!
    -0.14
    roscope
    -0.14
    verb
    -0.14
    lij
    -0.14
     Bij
    -0.14
    HU
    -0.14
    ~-
    -0.14
    agnost
    -0.14
    POSITIVE LOGITS
    imi
    0.15
    ouston
    0.14
     Struct
    0.14
     Lau
    0.14
     Laur
    0.14
    اÙĦÙĩ
    0.14
    vr
    0.13
    isphere
    0.13
    onds
    0.13
    βο
    0.13
    Act Density 0.097%

    No Known Activations