INDEX
    Explanations

    writing, watching, reading reviews

    New Auto-Interp
    Negative Logits
     betrieb
    0.71
     engineered
    0.71
     motorized
    0.71
     nozzle
    0.70
    打造
    0.70
     coaxial
    0.68
     control
    0.68
     marketplaces
    0.68
     project
    0.68
     welded
    0.67
    POSITIVE LOGITS
    0.99
    Reading
    0.98
    reading
    0.95
    欣赏
    0.94
    阅读
    0.91
    Watching
    0.91
    观看
    0.91
    emotional
    0.90
     menonton
    0.89
    欣賞
    0.88
    Act Density 0.002%

    No Known Activations