Hubert Wu: 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
91%z 5.02 - 2
90.7%z 4.96 - 3
89.6%z 4.80 - 4
88.9%z 4.74 - 5
88.7%z 4.73 - 6
87.6%z 4.64 - 7
85%z 4.42 - 8
84.4%z 4.37 - 9
83.3%z 4.28 - 10
82.7%z 4.24 - 11
82.5%z 4.22 - 12
81.9%z 4.17 - 13
81.6%z 4.15 - 14
81.5%z 4.14 - 15
81.5%z 4.14 - 16
81.3%z 4.12 - 17
81.1%z 4.11 - 18
80.3%z 4.04 - 19
80.2%z 4.03 - 20
80.1%z 4.03 - 21
76.2%z 3.98 - 22
73%z 3.94 - 23
69%z 3.90 - 24
66.9%z 3.87 - 25
61.3%z 3.81 - 26
59%z 3.79 - 27
58.7%z 3.78 - 28
58.7%z 3.78 - 29
58.5%z 3.78 - 30
55.7%z 3.75 - 31
55.4%z 3.75 - 32
53.9%z 3.73 - 33
51.7%z 3.71 - 34
49.9%z 3.69 - 35
49.9%z 3.69 - 36
48.1%z 3.67 - 37
47.9%z 3.66 - 38
47.9%z 3.66 - 39
47.5%z 3.63 - 40
47.1%z 3.61 - 41
46.9%z 3.59 - 42
46.7%z 3.58 - 43
46.5%z 3.57 - 44
45.6%z 3.51 - 45
45.4%z 3.50 - 46
45.4%z 3.50 - 47
45.4%z 3.50 - 48
45.3%z 3.49 - 49
45%z 3.47 - 50
44.9%z 3.47