133 lines
9 KiB
Text
133 lines
9 KiB
Text
re-gate: 55 chunks × 2 judges = 109 new tasks (1 done); empty-final (экспорт-баг, вне скора): [(5, 0), (20, 0)]
|
||
1.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
1.1 gpt-5-mini fid=96 style=90 cat=False flags=0 $0.0018
|
||
1.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0702
|
||
1.2 gpt-5-mini fid=95 style=88 cat=False flags=1 $0.0023
|
||
1.2 kimi-k2.6 fid=84 style=88 cat=True flags=3 $0.0437
|
||
1.3 gpt-5-mini fid=100 style=95 cat=False flags=0 $0.0003
|
||
1.3 kimi-k2.6 fid=98 style=90 cat=False flags=0 $0.0048
|
||
1.4 gpt-5-mini fid=95 style=90 cat=False flags=0 $0.0021
|
||
1.4 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
1.5 gpt-5-mini fid=92 style=88 cat=False flags=1 $0.0033
|
||
1.5 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
2.0 gpt-5-mini fid=95 style=90 cat=False flags=0 $0.0019
|
||
2.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0699
|
||
2.1 gpt-5-mini fid=95 style=85 cat=False flags=1 $0.0033
|
||
2.1 kimi-k2.6 fid=87 style=92 cat=False flags=3 $0.0676
|
||
3.0 gpt-5-mini fid=92 style=90 cat=False flags=1 $0.0025
|
||
3.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
3.1 gpt-5-mini fid=92 style=90 cat=False flags=1 $0.0021
|
||
3.1 kimi-k2.6 fid=85 style=78 cat=False flags=4 $0.0555
|
||
4.0 gpt-5-mini fid=95 style=90 cat=False flags=2 $0.0029
|
||
4.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0703
|
||
4.1 gpt-5-mini fid=95 style=90 cat=False flags=2 $0.0020
|
||
4.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0691
|
||
5.1 gpt-5-mini fid=92 style=90 cat=True flags=1 $0.0017
|
||
5.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
6.0 gpt-5-mini fid=98 style=92 cat=False flags=0 $0.0022
|
||
6.0 kimi-k2.6 fid=70 style=75 cat=True flags=2 $0.0668
|
||
6.1 gpt-5-mini fid=92 style=85 cat=False flags=2 $0.0025
|
||
6.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
7.0 gpt-5-mini fid=98 style=90 cat=False flags=0 $0.0021
|
||
7.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0702
|
||
7.1 gpt-5-mini fid=95 style=90 cat=False flags=2 $0.0024
|
||
7.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
8.0 gpt-5-mini fid=95 style=90 cat=False flags=1 $0.0028
|
||
8.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
8.1 gpt-5-mini fid=96 style=90 cat=False flags=1 $0.0021
|
||
8.1 kimi-k2.6 fid=92 style=90 cat=False flags=1 $0.0452
|
||
9.0 gpt-5-mini fid=94 style=92 cat=False flags=1 $0.0024
|
||
9.0 kimi-k2.6 fid=90 style=95 cat=False flags=3 $0.0352
|
||
9.1 gpt-5-mini fid=92 style=88 cat=False flags=2 $0.0035
|
||
9.1 kimi-k2.6 fid=90 style=93 cat=True flags=2 $0.0679
|
||
10.0 gpt-5-mini fid=95 style=90 cat=False flags=1 $0.0030
|
||
10.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0704
|
||
10.1 gpt-5-mini fid=60 style=90 cat=True flags=2 $0.0028
|
||
10.1 kimi-k2.6 fid=60 style=78 cat=True flags=3 $0.0273
|
||
11.0 gpt-5-mini fid=98 style=90 cat=False flags=0 $0.0016
|
||
11.0 kimi-k2.6 fid=86 style=None cat=True flags=0 $0.0688
|
||
11.1 gpt-5-mini fid=95 style=90 cat=False flags=1 $0.0022
|
||
11.1 kimi-k2.6 fid=95 style=93 cat=False flags=2 $0.0344
|
||
12.0 gpt-5-mini fid=98 style=90 cat=False flags=1 $0.0022
|
||
12.0 kimi-k2.6 fid=86 style=91 cat=True flags=5 $0.0498
|
||
12.1 gpt-5-mini fid=97 style=90 cat=False flags=1 $0.0022
|
||
12.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
13.0 gpt-5-mini fid=98 style=92 cat=False flags=0 $0.0022
|
||
13.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
13.1 gpt-5-mini fid=88 style=90 cat=False flags=3 $0.0028
|
||
13.1 kimi-k2.6 fid=90 style=86 cat=True flags=1 $0.0533
|
||
14.0 gpt-5-mini fid=92 style=90 cat=False flags=2 $0.0028
|
||
14.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0699
|
||
14.1 gpt-5-mini fid=93 style=85 cat=False flags=2 $0.0028
|
||
14.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
15.0 gpt-5-mini fid=97 style=92 cat=False flags=1 $0.0020
|
||
15.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
15.1 gpt-5-mini fid=78 style=90 cat=True flags=1 $0.0018
|
||
15.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
16.0 gpt-5-mini fid=90 style=88 cat=False flags=1 $0.0025
|
||
16.0 kimi-k2.6 fid=68 style=76 cat=True flags=4 $0.0686
|
||
16.1 gpt-5-mini fid=92 style=90 cat=False flags=1 $0.0023
|
||
16.1 kimi-k2.6 fid=88 style=84 cat=False flags=4 $0.0378
|
||
17.0 gpt-5-mini fid=95 style=90 cat=False flags=1 $0.0027
|
||
17.0 kimi-k2.6 fid=95 style=90 cat=False flags=3 $0.0663
|
||
17.1 gpt-5-mini fid=92 style=90 cat=False flags=2 $0.0029
|
||
17.1 kimi-k2.6 fid=92 style=90 cat=False flags=4 $0.0638
|
||
18.0 gpt-5-mini fid=95 style=92 cat=False flags=1 $0.0026
|
||
18.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
18.1 gpt-5-mini fid=90 style=85 cat=False flags=2 $0.0025
|
||
18.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
18.2 gpt-5-mini fid=97 style=92 cat=False flags=0 $0.0023
|
||
18.2 kimi-k2.6 fid=90 style=85 cat=False flags=3 $0.0501
|
||
19.0 gpt-5-mini fid=98 style=92 cat=False flags=0 $0.0021
|
||
19.0 kimi-k2.6 fid=96 style=92 cat=False flags=3 $0.0335
|
||
19.1 gpt-5-mini fid=96 style=92 cat=False flags=0 $0.0012
|
||
19.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
20.1 gpt-5-mini fid=96 style=88 cat=False flags=0 $0.0018
|
||
20.1 kimi-k2.6 fid=72 style=88 cat=True flags=3 $0.0498
|
||
21.0 gpt-5-mini fid=96 style=85 cat=False flags=1 $0.0025
|
||
21.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
21.1 gpt-5-mini fid=98 style=95 cat=False flags=0 $0.0019
|
||
21.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
22.0 gpt-5-mini fid=98 style=90 cat=False flags=2 $0.0027
|
||
22.0 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0700
|
||
22.1 gpt-5-mini fid=75 style=90 cat=False flags=3 $0.0035
|
||
22.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
22.2 gpt-5-mini fid=95 style=90 cat=False flags=1 $0.0015
|
||
22.2 kimi-k2.6 fid=93 style=90 cat=False flags=1 $0.0420
|
||
23.0 gpt-5-mini fid=98 style=92 cat=False flags=1 $0.0024
|
||
23.0 kimi-k2.6 fid=98 style=95 cat=False flags=1 $0.0295
|
||
23.1 gpt-5-mini fid=98 style=90 cat=False flags=0 $0.0031
|
||
23.1 kimi-k2.6 fid=85 style=87 cat=False flags=2 $0.0684
|
||
23.2 gpt-5-mini fid=92 style=85 cat=False flags=1 $0.0013
|
||
23.2 kimi-k2.6 fid=100 style=88 cat=False flags=0 $0.0222
|
||
24.0 gpt-5-mini fid=98 style=92 cat=False flags=0 $0.0020
|
||
24.0 kimi-k2.6 fid=93 style=88 cat=False flags=2 $0.0625
|
||
24.1 gpt-5-mini fid=95 style=92 cat=False flags=1 $0.0024
|
||
24.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0691
|
||
25.0 gpt-5-mini fid=96 style=88 cat=False flags=0 $0.0023
|
||
25.0 kimi-k2.6 fid=95 style=90 cat=True flags=2 $0.0466
|
||
25.1 gpt-5-mini fid=98 style=90 cat=False flags=0 $0.0014
|
||
25.1 kimi-k2.6 fid=None style=None cat=None flags=0 $0.0000
|
||
|
||
=== FIDELITY RE-GATE свод ===
|
||
gpt-5-mini: mean fidelity 93.7 (n=55, min=60)
|
||
kimi-k2.6: mean fidelity 87.6 (n=26, min=60)
|
||
leave-one-out: без gpt-5-mini → 87.6; без kimi-k2.6 → 93.7
|
||
КАТАСТРОФ-флаги: 13 → [('1.2', 'kimi-k2.6'), ('5.1', 'gpt-5-mini'), ('6.0', 'kimi-k2.6'), ('9.1', 'kimi-k2.6'), ('10.1', 'gpt-5-mini'), ('10.1', 'kimi-k2.6'), ('11.0', 'kimi-k2.6'), ('12.0', 'kimi-k2.6'), ('13.1', 'kimi-k2.6'), ('15.1', 'gpt-5-mini'), ('16.0', 'kimi-k2.6'), ('20.1', 'kimi-k2.6'), ('25.0', 'kimi-k2.6')]
|
||
ИНВЕРСИИ РЕДАКТУРЫ: 21
|
||
1.2 [gpt-5-mini]: Всё вышесказанное — предисловие к этой книге, оно же — преду
|
||
1.2 [kimi-k2.6]: вынуждать их отступать
|
||
5.1 [gpt-5-mini]: Душа расслабилась
|
||
6.0 [kimi-k2.6]: отозвался он
|
||
9.0 [kimi-k2.6]: тосковал по семейному уюту
|
||
9.1 [kimi-k2.6]: Фан Юань всего лишь разряда Б
|
||
10.1 [kimi-k2.6]: сосредоточить усилия на поисках
|
||
12.0 [kimi-k2.6]: Фан Юань насторожился
|
||
13.1 [gpt-5-mini]: Фан Юань с самодовольной усмешкой отвёл взгляд
|
||
13.1 [kimi-k2.6]: с самодовельной усмешкой
|
||
14.0 [gpt-5-mini]: останкам Монаха Цветочного Вина
|
||
15.1 [gpt-5-mini]: Хоть и всего лишь первый ранг
|
||
16.0 [kimi-k2.6]: Четвёртый старейшина рода
|
||
16.1 [gpt-5-mini]: Пусть результат и не вполне оправдал ожидания, он всё же был
|
||
16.1 [kimi-k2.6]: был предсказуем
|
||
экспорт-баг пустые финалы (вне скора): [(5, 0), (20, 0)]
|