Filip Petrušev: 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
82.3%z 4.21 - 2
75%z 3.96 - 3
72.1%z 3.93 - 4
60.9%z 3.81 - 5
47.6%z 3.64 - 6
47.2%z 3.62 - 7
47.1%z 3.61 - 8
43.4%z 3.37 - 9
43.3%z 3.36 - 10
42.6%z 3.31 - 11
42.5%z 3.31 - 12
42.3%z 3.30 - 13
42.1%z 3.28 - 14
42%z 3.28 - 15
41.5%z 3.24 - 16
41.5%z 3.24 - 17
41%z 3.21 - 18
40.1%z 3.16 - 19
40%z 3.15 - 20
39.9%z 3.14 - 21
39.9%z 3.14 - 22
39.9%z 3.14 - 23
39.6%z 3.12 - 24
39.1%z 3.09 - 25
38.6%z 3.06 - 26
38.5%z 3.05 - 27
37.9%z 3.01 - 28
37.8%z 3.01 - 29
37.7%z 3.00 - 30
37.6%z 2.99 - 31
37.4%z 2.98 - 32
37.2%z 2.97 - 33
37.1%z 2.96 - 34
37%z 2.96 - 35
37%z 2.95 - 36
36.9%z 2.95 - 37
36.9%z 2.94 - 38
36.7%z 2.94 - 39
36.5%z 2.92 - 40
36.4%z 2.92 - 41
36.3%z 2.91 - 42
36.2%z 2.90 - 43
36.1%z 2.90 - 44
36%z 2.89 - 45
36%z 2.89 - 46
35.9%z 2.88 - 47
35.8%z 2.88 - 48
35.8%z 2.88 - 49
35.6%z 2.87 - 50
35.6%z 2.86