Karena Lam: raw computed results
The raw ranking from the production similarity calculation (face-embedding centroid cosine → gender-cohort z-normalization). The web-opinion crown, curation, and fusion re-rank are removed.
- 1
94.3%z 5.64 - 2
92.5%z 5.30 - 3
90.4%z 4.92 - 4
89.1%z 4.76 - 5
86.5%z 4.55 - 6
86.3%z 4.53 - 7
83.6%z 4.31 - 8
82.7%z 4.23 - 9
82.6%z 4.23 - 10
82.2%z 4.20 - 11
80.1%z 4.03 - 12
76.8%z 3.98 - 13
76.5%z 3.98 - 14
71.5%z 3.92 - 15
70.5%z 3.91 - 16
70.3%z 3.91 - 17
65.2%z 3.85 - 18
63.9%z 3.84 - 19
63%z 3.83 - 20
62.1%z 3.82 - 21
61.8%z 3.82 - 22
61.5%z 3.81 - 23
58.5%z 3.78 - 24
55.7%z 3.75 - 25
55.6%z 3.75 - 26
55.5%z 3.75 - 27
53.6%z 3.73 - 28
50.9%z 3.70 - 29
50%z 3.69 - 30
49.3%z 3.68 - 31
48.3%z 3.67 - 32
48.2%z 3.67 - 33
47.6%z 3.64 - 34
47.6%z 3.64 - 35
47.2%z 3.61 - 36
47%z 3.60 - 37
46.9%z 3.60 - 38
46.8%z 3.59 - 39
46.5%z 3.57 - 40
46.4%z 3.57 - 41
46.2%z 3.55 - 42
46.1%z 3.55 - 43
45.9%z 3.53 - 44
45.9%z 3.53 - 45
45.8%z 3.52 - 46
45.6%z 3.51 - 47
45.6%z 3.51 - 48
45.5%z 3.50 - 49
45.1%z 3.48 - 50
45.1%z 3.48