INDEX
    Explanations

    repeated phrases or patterns related to "of" and "the"

    New Auto-Interp
    Negative Logits
     widow
    -0.15
    reuse
    -0.15
     Wid
    -0.15
    ulle
    -0.15
     ;
    -0.15
    ži
    -0.15
    ugu
    -0.14
     noises
    -0.14
    emo
    -0.14
     is
    -0.14
    POSITIVE LOGITS
    entine
    0.16
    apest
    0.16
    ovaly
    0.15
     MetroFramework
    0.15
    essional
    0.14
     Vance
    0.14
    leÅŁik
    0.14
    etas
    0.14
    isin
    0.14
    еÑĨÑĮ
    0.14
    Act Density 0.088%

    No Known Activations