INDEX
    Explanations

    contributes to harm normalization

    New Auto-Interp
    Negative Logits
    關鍵
    0.44
     incorporated
    0.39
    を切
    0.39
    യിലെ
    0.37
     spiega
    0.36
    一定要
    0.36
    截至
    0.36
     veloce
    0.35
     규칙
    0.35
     Curr
    0.35
    POSITIVE LOGITS
     perpet
    1.03
     perpetuate
    0.99
     harm
    0.89
     contributes
    0.89
     contribute
    0.87
     undermines
    0.82
    harm
    0.81
     potencialmente
    0.80
     demoral
    0.80
     dehuman
    0.79
    Act Density 0.110%

    No Known Activations