INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     Compute
    0.45
     Theoretical
    0.43
     substrate
    0.43
    Lecture
    0.42
     computed
    0.42
     ریاضی
    0.41
     Lecture
    0.41
     Software
    0.40
    规划
    0.40
     Enzymes
    0.40
    POSITIVE LOGITS
     news
    1.74
    報道
    1.68
     journalistic
    1.64
     headlines
    1.62
    报道
    1.62
    news
    1.61
     berita
    1.59
     journalists
    1.58
     journalist
    1.53
    新闻
    1.53
    Act Density 0.041%

    No Known Activations