Fernando: 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
81.8%z 4.17 - 2
80.6%z 4.06 - 3
61.3%z 3.81 - 4
57.8%z 3.77 - 5
47.6%z 3.64 - 6
47.5%z 3.63 - 7
47%z 3.60 - 8
43.9%z 3.40 - 9
43.7%z 3.38 - 10
43.5%z 3.37 - 11
43.4%z 3.37 - 12
43%z 3.34 - 13
42.9%z 3.34 - 14
42.9%z 3.33 - 15
42.5%z 3.31 - 16
42.5%z 3.31 - 17
42.2%z 3.29 - 18
42%z 3.28 - 19
41.7%z 3.26 - 20
41.5%z 3.25 - 21
41.2%z 3.23 - 22
41.2%z 3.22 - 23
40.7%z 3.19 - 24
40.3%z 3.17 - 25
40.2%z 3.16 - 26
40.2%z 3.16 - 27
40.1%z 3.15 - 28
40%z 3.15 - 29
39.9%z 3.14 - 30
39.9%z 3.14 - 31
39.8%z 3.14 - 32
39.7%z 3.13 - 33
39.5%z 3.12 - 34
39.1%z 3.09 - 35
39.1%z 3.09 - 36
38.9%z 3.08 - 37
38.9%z 3.07 - 38
38.8%z 3.07 - 39
38.5%z 3.05 - 40
37.9%z 3.01 - 41
37.9%z 3.01 - 42
37.8%z 3.01 - 43
37.7%z 3.00 - 44
37.5%z 2.99 - 45
37.4%z 2.98 - 46
37.4%z 2.98 - 47
37.4%z 2.98 - 48
37.3%z 2.98 - 49
37.3%z 2.97 - 50
37.3%z 2.97