| open jev / ir control v1 |
Noul |
100.0 |
100.0 |
60.3 |
|
来源 ↗ |
| open jev / mailroom control v1 |
Choice |
100.0 |
100.0 |
89.6 |
|
来源 ↗ |
| open jev / sponsor segment control v1 |
Choice |
100.0 |
100.0 |
66.4 |
|
来源 ↗ |
| open jev / vizdoom basic v1 |
Noul |
100.0 |
100.0 |
53.2 |
|
来源 ↗ |
| generalization japanese diverse |
Choice |
99.5 |
100.0 |
94.2 |
|
来源 ↗ |
| generalization japanese diverse |
Noul |
99.3 |
100.0 |
91.4 |
|
来源 ↗ |
| s1mb generalization contextual choice |
Choice |
98.7 |
100.0 |
88.7 |
|
来源 ↗ |
| open jev / browser control v1 |
Choice |
97.8 |
100.0 |
76.5 |
|
来源 ↗ |
| dbpedia |
Choice |
97.8 |
98.9 |
88.9 |
|
来源 ↗ |
| open jev / citation control v1 |
Choice |
97.0 |
100.0 |
55.9 |
|
来源 ↗ |
| laya / enron spam |
Noul |
96.0 |
100.0 |
64.9 |
|
来源 ↗ |
| jmteb retrieval |
Choice |
94.9 |
98.7 |
90.0 |
|
来源 ↗ |
| generalization japanese knowledge hard |
Choice |
94.2 |
99.6 |
88.0 |
|
来源 ↗ |
| jcommonsenseqa |
Choice |
93.8 |
97.9 |
87.6 |
|
来源 ↗ |
| s1mb generalization diverse noul |
Noul |
92.0 |
100.0 |
70.7 |
|
来源 ↗ |
| s1mb generalization contextual noul |
Noul |
90.0 |
100.0 |
72.1 |
|
来源 ↗ |
| open jev / email selection control v1 |
Choice |
89.6 |
100.0 |
38.4 |
|
来源 ↗ |
| open jev / painting geometry v1 |
Choice |
89.5 |
100.0 |
60.5 |
|
来源 ↗ |
| janli |
Choice |
88.9 |
88.9 |
41.7 |
|
来源 ↗ |
| miqa |
Choice |
88.6 |
97.7 |
71.1 |
|
来源 ↗ |
| open jev / drone control v1 |
Noul |
88.2 |
100.0 |
63.8 |
|
来源 ↗ |
| s1mb generalization diverse choice |
Choice |
88.1 |
100.0 |
78.1 |
|
来源 ↗ |
| commonsensemoralja |
Noul |
88.0 |
89.3 |
67.8 |
|
来源 ↗ |
| sdoh nli |
Noul |
88.0 |
100.0 |
70.0 |
|
来源 ↗ |
| arc |
Choice |
86.8 |
98.5 |
73.8 |
|
来源 ↗ |
| bbq |
Choice |
86.7 |
98.3 |
64.5 |
|
来源 ↗ |
| openbookqa |
Choice |
86.1 |
98.6 |
64.4 |
|
来源 ↗ |
| argument quality |
Choice |
85.7 |
95.9 |
53.3 |
|
来源 ↗ |
| massive |
Choice |
84.6 |
97.8 |
73.4 |
|
来源 ↗ |
| open jev / email selection control v1 |
Noul |
84.6 |
100.0 |
56.9 |
|
来源 ↗ |
| open jev / amount extraction control v1 |
Noul |
84.4 |
100.0 |
44.2 |
|
来源 ↗ |
| open jev / amount extraction control v1 |
Choice |
84.3 |
100.0 |
63.2 |
|
来源 ↗ |
| laya / ag news |
Choice |
83.6 |
94.0 |
77.7 |
|
来源 ↗ |
| ethics |
Choice |
83.3 |
100.0 |
28.5 |
|
来源 ↗ |
| open jev / workflow controls v1 / invoice processing |
Noul |
82.7 |
100.0 |
47.0 |
|
来源 ↗ |
| open jev / customer control v1 |
Choice |
82.6 |
82.6 |
71.5 |
|
来源 ↗ |
| open jev / customer control v1 |
Noul |
82.3 |
100.0 |
68.3 |
|
来源 ↗ |
| snli |
Choice |
82.3 |
98.4 |
59.9 |
|
来源 ↗ |
| hans |
Noul |
82.0 |
100.0 |
53.1 |
|
来源 ↗ |
| scitail |
Noul |
82.0 |
98.0 |
63.4 |
|
来源 ↗ |
| open jev / mailroom control v1 |
Noul |
81.9 |
100.0 |
73.2 |
|
来源 ↗ |
| open jev / phone extraction control v1 |
Noul |
81.8 |
100.0 |
63.9 |
|
来源 ↗ |
| jamp |
Choice |
81.5 |
81.5 |
50.0 |
|
来源 ↗ |
| canttalk |
Noul |
80.0 |
92.0 |
50.5 |
|
来源 ↗ |
| impli |
Noul |
80.0 |
96.0 |
57.1 |
|
来源 ↗ |
| quartz |
Choice |
78.6 |
90.5 |
59.4 |
|
来源 ↗ |
| open jev / silent failure control v1 |
Noul |
78.0 |
100.0 |
49.8 |
|
来源 ↗ |
| banking77 |
Choice |
77.9 |
93.7 |
64.2 |
|
来源 ↗ |
| hwu64 |
Choice |
77.7 |
95.7 |
67.3 |
|
来源 ↗ |
| open jev / phone extraction control v1 |
Choice |
76.1 |
100.0 |
44.4 |
|
来源 ↗ |
| hatecheck |
Noul |
76.0 |
100.0 |
57.5 |
|
来源 ↗ |
| kuci |
Choice |
76.0 |
85.8 |
64.3 |
|
来源 ↗ |
| s1mb generalization contextual score |
Score |
75.5 |
99.3 |
52.6 |
|
来源 ↗ |
| clinc |
Choice |
75.2 |
99.3 |
70.5 |
|
来源 ↗ |
| mtop |
Choice |
74.3 |
93.2 |
67.4 |
|
来源 ↗ |
| jmmlu japan |
Choice |
73.8 |
94.2 |
68.4 |
|
来源 ↗ |
| followir robust04 |
Noul |
72.3 |
90.9 |
53.7 |
|
来源 ↗ |
| gretel pii |
Choice |
72.1 |
94.8 |
51.1 |
|
来源 ↗ |
| open jev / ir control v1 |
Choice |
72.0 |
88.0 |
55.5 |
|
来源 ↗ |
| winobias |
Choice |
71.7 |
95.7 |
42.7 |
|
来源 ↗ |
| jnli |
Choice |
71.5 |
87.5 |
68.4 |
|
来源 ↗ |
| creak |
Noul |
70.0 |
90.0 |
48.2 |
|
来源 ↗ |
| open jev / context retention control v1 |
Noul |
70.0 |
100.0 |
35.2 |
|
来源 ↗ |
| arct |
Choice |
68.1 |
89.4 |
49.3 |
|
来源 ↗ |
| s1mb generalization diverse score |
Score |
67.5 |
97.6 |
52.7 |
|
来源 ↗ |
| ethos |
Noul |
66.0 |
76.0 |
49.5 |
|
来源 ↗ |
| paws |
Noul |
66.0 |
92.0 |
52.5 |
|
来源 ↗ |
| jsts |
Score |
65.8 |
69.9 |
55.3 |
|
来源 ↗ |
| lonli |
Choice |
65.6 |
98.4 |
52.8 |
|
来源 ↗ |
| open jev / workflow controls v1 / agent trace observability |
Noul |
65.4 |
100.0 |
39.4 |
|
来源 ↗ |
| scone |
Choice |
64.0 |
98.0 |
48.5 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonq |
Noul |
63.9 |
72.7 |
48.0 |
|
来源 ↗ |
| aegis2 |
Noul |
63.8 |
68.3 |
39.6 |
|
来源 ↗ |
| few nerd |
Choice |
63.1 |
87.8 |
52.9 |
|
来源 ↗ |
| defeasible nli |
Choice |
63.0 |
89.1 |
43.1 |
|
来源 ↗ |
| open jev / workflow controls v1 / customer service |
Noul |
62.8 |
100.0 |
53.7 |
|
来源 ↗ |
| ruletaker |
Noul |
62.0 |
100.0 |
39.3 |
|
来源 ↗ |
| laya / typed decisions |
Score |
61.9 |
76.0 |
35.4 |
|
来源 ↗ |
| open jev / entity alignment control v1 |
Noul |
60.7 |
100.0 |
50.6 |
|
来源 ↗ |
| babi nli |
Noul |
60.0 |
96.0 |
48.5 |
|
来源 ↗ |
| synthetic relevance / nanocoir |
Noul |
59.0 |
73.6 |
42.2 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanomsmarco |
Noul |
58.7 |
64.1 |
45.3 |
|
来源 ↗ |
| jsem |
Choice |
56.9 |
69.2 |
45.9 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanohotpotqa |
Noul |
56.6 |
78.9 |
36.1 |
|
来源 ↗ |
| scicite |
Choice |
56.5 |
74.2 |
49.4 |
|
来源 ↗ |
| temporal nli |
Choice |
56.5 |
100.0 |
28.0 |
|
来源 ↗ |
| generalization japanese knowledge hard |
Noul |
54.7 |
87.3 |
50.6 |
|
来源 ↗ |
| ethics |
Noul |
53.8 |
80.6 |
33.4 |
|
来源 ↗ |
| open jev / workflow controls v1 / security incidents |
Noul |
53.6 |
100.0 |
50.0 |
|
来源 ↗ |
| jcola |
Noul |
52.5 |
57.2 |
21.6 |
|
来源 ↗ |
| lexcomp |
Noul |
52.0 |
60.0 |
27.7 |
|
来源 ↗ |
| poem sentiment |
Choice |
51.4 |
68.6 |
9.3 |
|
来源 ↗ |
| qasper |
Noul |
50.0 |
88.0 |
51.4 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanofiqa2018 |
Noul |
49.8 |
58.0 |
35.3 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanodbpedia |
Noul |
49.3 |
53.2 |
33.8 |
|
来源 ↗ |
| open jev / reasoning control v1 |
Score |
48.8 |
96.3 |
38.3 |
|
来源 ↗ |
| jsick sts |
Score |
48.6 |
61.5 |
40.1 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoscidocs |
Noul |
48.5 |
48.5 |
21.5 |
|
来源 ↗ |
| contract nli |
Choice |
47.7 |
81.6 |
43.0 |
|
来源 ↗ |
| go emotions |
Noul |
47.0 |
63.4 |
42.1 |
|
来源 ↗ |
| mmmlu |
Choice |
46.7 |
78.2 |
52.8 |
|
来源 ↗ |
| ud ewt |
Choice |
46.6 |
96.0 |
19.6 |
|
来源 ↗ |
| open jev / ir control v1 |
Score |
45.4 |
100.0 |
23.0 |
|
来源 ↗ |
| open jev / painting geometry v1 |
Noul |
44.8 |
100.0 |
42.0 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonfcorpus |
Noul |
44.5 |
49.7 |
29.6 |
|
来源 ↗ |
| followir news21 |
Noul |
44.0 |
58.0 |
18.4 |
|
来源 ↗ |
| open jev / customer control v1 |
Score |
43.4 |
99.3 |
44.4 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanotouche2020 |
Noul |
43.3 |
48.2 |
32.3 |
|
来源 ↗ |
| wiqa |
Choice |
42.9 |
68.3 |
24.0 |
|
来源 ↗ |
| generalization japanese knowledge hard |
Score |
42.0 |
84.6 |
45.3 |
|
来源 ↗ |
| tracie |
Noul |
42.0 |
68.0 |
14.4 |
|
来源 ↗ |
| crows pairs |
Choice |
41.7 |
77.1 |
37.9 |
|
来源 ↗ |
| laya / typed decisions |
Noul |
41.1 |
49.2 |
29.9 |
|
来源 ↗ |
| cladder |
Noul |
40.0 |
94.0 |
28.5 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanohotpotqa |
Score |
38.7 |
71.8 |
20.9 |
|
来源 ↗ |
| winowhy |
Noul |
36.0 |
66.0 |
25.9 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanodbpedia |
Score |
35.4 |
50.3 |
23.4 |
|
来源 ↗ |
| patent similarity |
Score |
35.4 |
53.0 |
24.0 |
|
来源 ↗ |
| robust lr |
Choice |
35.0 |
98.3 |
29.3 |
|
来源 ↗ |
| civil comments |
Noul |
34.2 |
52.8 |
25.2 |
|
来源 ↗ |
| open jev / vizdoom basic v1 |
Score |
34.2 |
100.0 |
42.3 |
|
来源 ↗ |
| mintaka ja boolean |
Noul |
34.0 |
68.0 |
37.5 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanotouche2020 |
Score |
33.9 |
49.4 |
25.5 |
|
来源 ↗ |
| open jev / reasoning control v1 |
Noul |
33.5 |
100.0 |
34.3 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanomsmarco |
Score |
31.9 |
57.3 |
25.0 |
|
来源 ↗ |
| stepgame |
Choice |
30.9 |
82.1 |
24.8 |
|
来源 ↗ |
| jamc qa |
Choice |
30.2 |
65.3 |
23.0 |
|
来源 ↗ |
| laya / typed decisions |
Choice |
29.7 |
35.7 |
23.3 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoarguana |
Noul |
29.4 |
53.5 |
22.0 |
|
来源 ↗ |
| sobaco culture |
Choice |
28.5 |
80.5 |
30.6 |
|
来源 ↗ |
| followir core17 |
Noul |
27.9 |
46.6 |
25.2 |
|
来源 ↗ |
| sgd |
Choice |
27.1 |
96.6 |
28.6 |
|
来源 ↗ |
| open jev / reasoning control v1 |
Choice |
25.0 |
100.0 |
41.7 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoarguana |
Score |
24.9 |
43.2 |
10.3 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonfcorpus |
Score |
24.2 |
40.9 |
15.8 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanofiqa2018 |
Score |
23.9 |
49.0 |
21.0 |
|
来源 ↗ |
| boardgameqa |
Choice |
23.4 |
90.6 |
24.4 |
|
来源 ↗ |
| hh rlhf |
Choice |
23.1 |
33.3 |
4.5 |
|
来源 ↗ |
| open jev / snake v1 |
Noul |
22.3 |
100.0 |
19.1 |
|
来源 ↗ |
| mmlu prox ja |
Choice |
21.5 |
69.5 |
39.5 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonq |
Score |
20.4 |
54.3 |
13.3 |
|
来源 ↗ |
| plane |
Noul |
20.0 |
100.0 |
12.6 |
|
来源 ↗ |
| fol nli |
Choice |
19.7 |
75.4 |
17.3 |
|
来源 ↗ |
| logical entailment |
Noul |
18.0 |
84.0 |
18.7 |
|
来源 ↗ |
| gpqa ja main |
Choice |
16.4 |
54.7 |
20.9 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoscidocs |
Score |
15.8 |
36.7 |
10.7 |
|
来源 ↗ |
| argument quality |
Noul |
15.7 |
30.2 |
7.8 |
|
来源 ↗ |
| generalization japanese math formal hard |
Choice |
14.2 |
48.4 |
21.0 |
|
来源 ↗ |
| gsm8k |
Choice |
11.4 |
95.7 |
38.0 |
|
来源 ↗ |
| open jev / entity alignment control v1 |
Score |
10.6 |
100.0 |
18.4 |
|
来源 ↗ |
| nlsat |
Noul |
10.0 |
22.0 |
3.2 |
|
来源 ↗ |
| corr2cause |
Choice |
8.3 |
91.7 |
9.3 |
|
来源 ↗ |
| spartqa |
Choice |
8.1 |
93.5 |
14.0 |
|
来源 ↗ |
| open jev / drone control v1 |
Choice |
7.9 |
100.0 |
13.4 |
|
来源 ↗ |
| jculture mcq |
Noul |
7.6 |
44.3 |
12.8 |
|
来源 ↗ |
| esci |
Choice |
6.8 |
49.3 |
17.5 |
|
来源 ↗ |
| synthetic relevance / nanocoir |
Score |
6.3 |
43.8 |
6.8 |
|
来源 ↗ |
| generalization japanese math formal hard |
Noul |
2.7 |
34.7 |
10.6 |
|
来源 ↗ |
| aqua rat |
Choice |
0.0 |
83.1 |
19.7 |
|
来源 ↗ |
| generalization japanese math formal hard |
Score |
0.0 |
6.9 |
0.9 |
|
来源 ↗ |
| open jev / drone control v1 |
Score |
0.0 |
100.0 |
7.2 |
|
来源 ↗ |
| open jev / painting geometry v1 |
Score |
0.0 |
100.0 |
29.7 |
|
来源 ↗ |
| open jev / snake v1 |
Choice |
0.0 |
83.5 |
18.3 |
|
来源 ↗ |
| open jev / tic tac toe v1 |
Choice |
0.0 |
54.5 |
5.1 |
|
来源 ↗ |