Peter S. Kim: 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
94.7%z 5.72 - 2
93.2%z 5.43 - 3
91.6%z 5.14 - 4
87.6%z 4.64 - 5
86.5%z 4.55 - 6
86%z 4.51 - 7
85.7%z 4.49 - 8
82.1%z 4.19 - 9
81.5%z 4.14 - 10
81.1%z 4.10 - 11
80.3%z 4.04 - 12
73.9%z 3.95 - 13
72.9%z 3.94 - 14
70%z 3.91 - 15
69.5%z 3.90 - 16
65%z 3.85 - 17
56.8%z 3.76 - 18
56.2%z 3.76 - 19
55.4%z 3.75 - 20
54.4%z 3.74 - 21
51.2%z 3.70 - 22
50.1%z 3.69 - 23
47.8%z 3.65 - 24
47.5%z 3.64 - 25
47.5%z 3.64 - 26
47.5%z 3.63 - 27
47.3%z 3.62 - 28
47%z 3.60 - 29
46.8%z 3.59 - 30
46.8%z 3.59 - 31
46.6%z 3.57 - 32
45.9%z 3.53 - 33
45.7%z 3.52 - 34
45.6%z 3.51 - 35
45.6%z 3.51 - 36
45.6%z 3.51 - 37
45.5%z 3.50 - 38
45.4%z 3.50 - 39
45.2%z 3.48 - 40
45.1%z 3.48 - 41
45.1%z 3.48 - 42
45%z 3.47 - 43
44.8%z 3.46 - 44
44.7%z 3.45 - 45
44.6%z 3.45 - 46
44.6%z 3.44 - 47
44%z 3.41 - 48
43.8%z 3.40 - 49
43.6%z 3.38 - 50
43.5%z 3.38