Mark Liu: 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
85.5%z 4.46 - 3
84.6%z 4.39 - 4
82.7%z 4.24 - 5
82.2%z 4.20 - 6
81.6%z 4.15 - 7
81.2%z 4.12 - 8
80.3%z 4.04 - 9
80.2%z 4.03 - 10
74.4%z 3.96 - 11
74.1%z 3.95 - 12
67.8%z 3.88 - 13
65%z 3.85 - 14
62.7%z 3.83 - 15
53.2%z 3.72 - 16
51.4%z 3.70 - 17
47.2%z 3.61 - 18
46.9%z 3.59 - 19
46.9%z 3.59 - 20
46.2%z 3.55 - 21
46.2%z 3.55 - 22
46%z 3.54 - 23
46%z 3.53 - 24
45.9%z 3.53 - 25
45.9%z 3.53 - 26
45.6%z 3.51 - 27
44.9%z 3.47 - 28
44.6%z 3.45 - 29
44.6%z 3.45 - 30
44.4%z 3.43 - 31
44.3%z 3.43 - 32
43.7%z 3.39 - 33
43.3%z 3.36 - 34
43.2%z 3.36 - 35
43.2%z 3.36 - 36
43.1%z 3.35 - 37
43.1%z 3.35 - 38
43%z 3.34 - 39
42.7%z 3.33 - 40
42.7%z 3.32 - 41
42.6%z 3.32 - 42
42.6%z 3.31 - 43
42.5%z 3.31 - 44
42.5%z 3.31 - 45
42.1%z 3.28 - 46
42.1%z 3.28 - 47
42%z 3.28 - 48
42%z 3.28 - 49
41.9%z 3.27 - 50
41.9%z 3.27