INDEX
    Explanations

    phrases indicating a desire or request

    expressions of desire or intent

    New Auto-Interp
    Negative Logits
     correct
    -0.69
     forth
    -0.65
     reg
    -0.64
     seals
    -0.63
     fragments
    -0.63
     consistent
    -0.63
     borne
    -0.63
     hy
    -0.62
     stag
    -0.62
     propag
    -0.62
    POSITIVE LOGITS
    Want
    3.11
     Want
    3.11
    want
    1.68
     WANT
    1.56
     Need
    1.33
    Interested
    1.25
     Wanted
    1.25
    Need
    1.16
     want
    1.11
     Feel
    1.11
    Act Density 0.013%

    No Known Activations