INDEX
    Explanations

    the word "to" with varying strength

    instances of the phrase "it’s hard to"

    New Auto-Interp
    Negative Logits
     Conditions
    -0.72
    plates
    -0.68
     embodiments
    -0.66
     Flags
    -0.65
    items
    -0.64
    cand
    -0.60
     Votes
    -0.60
    faces
    -0.59
    Maker
    -0.58
     Sierra
    -0.58
    POSITIVE LOGITS
     imagine
    1.34
     reconcile
    1.30
     quantify
    1.25
     distinguish
    1.22
     justify
    1.20
     convince
    1.17
     conceive
    1.17
     discern
    1.15
     differentiate
    1.14
     compare
    1.13
    Act Density 0.074%

    No Known Activations