Daniel Gros: 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
89.7%z 4.81 - 2
88.1%z 4.68 - 3
85.4%z 4.46 - 4
75.8%z 3.97 - 5
72.6%z 3.94 - 6
65.6%z 3.86 - 7
64.5%z 3.85 - 8
59%z 3.79 - 9
56%z 3.75 - 10
55.7%z 3.75 - 11
54.7%z 3.74 - 12
52%z 3.71 - 13
51.4%z 3.70 - 14
49.5%z 3.68 - 15
49%z 3.68 - 16
47.9%z 3.66 - 17
47.7%z 3.65 - 18
47.7%z 3.64 - 19
47.1%z 3.61 - 20
46.9%z 3.60 - 21
46.9%z 3.60 - 22
46.6%z 3.58 - 23
46.4%z 3.57 - 24
46.2%z 3.55 - 25
46.1%z 3.54 - 26
46%z 3.54 - 27
45.9%z 3.53 - 28
45%z 3.47 - 29
44.2%z 3.42 - 30
44%z 3.41 - 31
43.3%z 3.36 - 32
43.1%z 3.35 - 33
43.1%z 3.35 - 34
42.7%z 3.33 - 35
42.5%z 3.31 - 36
42.4%z 3.30 - 37
42.3%z 3.30 - 38
42.1%z 3.28 - 39
42%z 3.28 - 40
41.8%z 3.26 - 41
41.4%z 3.24 - 42
41.4%z 3.24 - 43
41.3%z 3.23 - 44
40.6%z 3.19 - 45
40.5%z 3.18 - 46
40.3%z 3.17 - 47
39.9%z 3.14 - 48
39.9%z 3.14 - 49
39.8%z 3.14 - 50
39.5%z 3.12