INDEX
    Explanations

    extract phrases from video

    New Auto-Interp
    Negative Logits
    an
    2.47
    ان
    2.36
    is
    2.34
    ak
    2.27
    ar
    2.23
    2.22
    in
    2.17
    al
    2.17
    el
    2.06
    u
    2.06
    POSITIVE LOGITS
    ς
    1.70
     the
    1.46
    sail
    1.44
    Н
    1.44
    szer
    1.38
     
    1.38
     an
    1.37
    และ
    1.34
    ТА
    1.34
    𝐬
    1.33
    Act Density 0.914%

    No Known Activations