Michel Lambert: 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
61.5%z 3.81 - 2
47.4%z 3.63 - 3
46.1%z 3.55 - 4
45.4%z 3.50 - 5
45.2%z 3.48 - 6
45.1%z 3.48 - 7
43.9%z 3.40 - 8
43.8%z 3.39 - 9
43.4%z 3.37 - 10
43.3%z 3.36 - 11
42.9%z 3.34 - 12
42.4%z 3.31 - 13
41.6%z 3.25 - 14
41.2%z 3.22 - 15
40.3%z 3.17 - 16
40%z 3.15 - 17
39.1%z 3.09 - 18
38.7%z 3.06 - 19
38.6%z 3.06 - 20
38.4%z 3.04 - 21
38.2%z 3.03 - 22
37.8%z 3.01 - 23
37.4%z 2.98 - 24
36.7%z 2.94 - 25
36.7%z 2.93 - 26
36.6%z 2.93 - 27
36.6%z 2.92 - 28
36.4%z 2.91 - 29
36%z 2.89 - 30
35.9%z 2.88 - 31
35.7%z 2.87 - 32
35.6%z 2.86 - 33
35.5%z 2.86 - 34
35.3%z 2.84 - 35
34.4%z 2.78 - 36
34.4%z 2.78 - 37
34.4%z 2.78 - 38
34.3%z 2.78 - 39
34.1%z 2.77 - 40
33.9%z 2.75 - 41
33.9%z 2.75 - 42
33.5%z 2.72 - 43
33.1%z 2.70 - 44
33.1%z 2.70 - 45
32.9%z 2.69 - 46
32.6%z 2.67 - 47
32.6%z 2.67 - 48
32.5%z 2.66 - 49
32.5%z 2.66 - 50
32.5%z 2.66