Tom 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
99%z 6.79 - 2
96.1%z 5.99 - 3
92.4%z 5.28 - 4
82.4%z 4.22 - 5
81.7%z 4.16 - 6
79.9%z 4.02 - 7
77.8%z 3.99 - 8
69.8%z 3.91 - 9
65.7%z 3.86 - 10
51.4%z 3.70 - 11
47.9%z 3.66 - 12
46%z 3.53 - 13
44.7%z 3.45 - 14
44.4%z 3.43 - 15
43.9%z 3.40 - 16
43.5%z 3.38 - 17
43.4%z 3.37 - 18
43.4%z 3.37 - 19
43.3%z 3.36 - 20
42.5%z 3.31 - 21
42.4%z 3.30 - 22
42.3%z 3.30 - 23
41.9%z 3.27 - 24
41.4%z 3.24 - 25
41.1%z 3.22 - 26
41%z 3.21 - 27
41%z 3.21 - 28
40.9%z 3.20 - 29
40.8%z 3.20 - 30
40.7%z 3.19 - 31
40%z 3.15 - 32
39.5%z 3.11 - 33
39.4%z 3.11 - 34
39.4%z 3.11 - 35
39.4%z 3.11 - 36
39.3%z 3.10 - 37
39.3%z 3.10 - 38
39.2%z 3.10 - 39
39.1%z 3.09 - 40
38.7%z 3.06 - 41
38.4%z 3.05 - 42
38.3%z 3.04 - 43
38.3%z 3.04 - 44
38.2%z 3.03 - 45
38.1%z 3.03 - 46
38%z 3.02 - 47
37.9%z 3.01 - 48
37.8%z 3.00 - 49
37.7%z 3.00 - 50
37.6%z 2.99