Mark Dragunski: 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
83.2%z 4.28 - 2
81.7%z 4.15 - 3
80.5%z 4.06 - 4
64.1%z 3.84 - 5
62%z 3.82 - 6
61.4%z 3.81 - 7
53.1%z 3.72 - 8
48%z 3.67 - 9
46.9%z 3.59 - 10
46.7%z 3.58 - 11
46.5%z 3.57 - 12
46.4%z 3.56 - 13
45.9%z 3.53 - 14
44.8%z 3.46 - 15
44.8%z 3.46 - 16
44.4%z 3.43 - 17
44.3%z 3.43 - 18
43.9%z 3.40 - 19
43.4%z 3.37 - 20
42%z 3.28 - 21
41.1%z 3.22 - 22
40.8%z 3.20 - 23
40.6%z 3.19 - 24
40.4%z 3.17 - 25
40.4%z 3.17 - 26
39.8%z 3.14 - 27
39.7%z 3.13 - 28
39.7%z 3.13 - 29
39.5%z 3.12 - 30
39%z 3.08 - 31
38.4%z 3.04 - 32
38.3%z 3.04 - 33
38.2%z 3.03 - 34
38.1%z 3.02 - 35
38%z 3.02 - 36
37.9%z 3.01 - 37
37.8%z 3.01 - 38
37.8%z 3.01 - 39
37.8%z 3.00 - 40
37.5%z 2.99 - 41
37.4%z 2.98 - 42
37.3%z 2.97 - 43
37.1%z 2.96 - 44
37%z 2.96 - 45
37%z 2.95 - 46
36.9%z 2.95 - 47
36.9%z 2.95 - 48
36.9%z 2.95 - 49
36.9%z 2.95 - 50
36.5%z 2.92