Filippo Berardi: 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
81.9%z 4.17 - 2
62.5%z 3.83 - 3
57.9%z 3.78 - 4
48%z 3.66 - 5
46.6%z 3.57 - 6
45.6%z 3.51 - 7
45.4%z 3.50 - 8
45.2%z 3.49 - 9
44.7%z 3.45 - 10
44.6%z 3.45 - 11
44.3%z 3.43 - 12
44.3%z 3.43 - 13
43.8%z 3.39 - 14
43.5%z 3.38 - 15
41.1%z 3.22 - 16
40.4%z 3.17 - 17
39.3%z 3.10 - 18
39.3%z 3.10 - 19
39.1%z 3.09 - 20
38.7%z 3.06 - 21
38.5%z 3.05 - 22
38.3%z 3.04 - 23
37%z 2.95 - 24
36.8%z 2.94 - 25
36.5%z 2.92 - 26
36.5%z 2.92 - 27
36.5%z 2.92 - 28
36.1%z 2.90 - 29
34.7%z 2.81 - 30
34.6%z 2.80 - 31
34.5%z 2.79 - 32
34.4%z 2.79 - 33
34.1%z 2.77 - 34
34.1%z 2.77 - 35
33.9%z 2.75 - 36
33.8%z 2.74 - 37
33.5%z 2.73 - 38
33.4%z 2.72 - 39
33.2%z 2.71 - 40
33%z 2.69 - 41
32.9%z 2.69 - 42
32.8%z 2.68 - 43
32.7%z 2.68 - 44
32.3%z 2.65 - 45
32.3%z 2.65 - 46
32.3%z 2.65 - 47
32.3%z 2.65 - 48
32.2%z 2.64 - 49
32.2%z 2.64