INDEX
    Explanations
    New Auto-Interp
    Negative Logits
    which
    0.67
    があった
    0.60
     contains
    0.56
     contain
    0.56
    I
    0.55
     والتي
    0.53
     które
    0.53
    containing
    0.52
    があれば
    0.51
     которое
    0.50
    POSITIVE LOGITS
     trying
    1.46
     afraid
    1.29
     interested
    1.26
    trying
    1.20
     aware
    1.18
     struggling
    1.18
     attempting
    1.15
     obsessed
    1.14
     willing
    1.11
     worried
    1.10
    Act Density 0.627%

    No Known Activations