| open jev / browser control v1 |
Choice |
100.0 |
100.0 |
68.3 |
|
来源 ↗ |
| open jev / citation control v1 |
Choice |
100.0 |
100.0 |
43.6 |
|
来源 ↗ |
| open jev / customer control v1 |
Noul |
100.0 |
100.0 |
59.1 |
|
来源 ↗ |
| open jev / drone control v1 |
Noul |
100.0 |
100.0 |
52.5 |
|
来源 ↗ |
| open jev / email selection control v1 |
Choice |
100.0 |
100.0 |
26.3 |
|
来源 ↗ |
| open jev / email selection control v1 |
Noul |
100.0 |
100.0 |
43.5 |
|
来源 ↗ |
| open jev / entity alignment control v1 |
Noul |
100.0 |
100.0 |
38.9 |
|
来源 ↗ |
| open jev / ir control v1 |
Noul |
100.0 |
100.0 |
49.1 |
|
来源 ↗ |
| open jev / mailroom control v1 |
Choice |
100.0 |
100.0 |
83.8 |
|
来源 ↗ |
| open jev / phone extraction control v1 |
Noul |
100.0 |
100.0 |
52.9 |
|
来源 ↗ |
| open jev / sponsor segment control v1 |
Choice |
100.0 |
100.0 |
54.3 |
|
来源 ↗ |
| open jev / vizdoom basic v1 |
Noul |
100.0 |
100.0 |
42.1 |
|
来源 ↗ |
| s1mb generalization contextual choice |
Choice |
98.7 |
100.0 |
83.3 |
|
来源 ↗ |
| open jev / phone extraction control v1 |
Choice |
97.8 |
100.0 |
34.9 |
|
来源 ↗ |
| dbpedia |
Choice |
97.8 |
98.9 |
84.2 |
|
来源 ↗ |
| arc |
Choice |
95.6 |
98.5 |
65.1 |
|
来源 ↗ |
| clinc |
Choice |
95.2 |
99.3 |
58.8 |
|
来源 ↗ |
| open jev / mailroom control v1 |
Noul |
95.0 |
100.0 |
64.0 |
|
来源 ↗ |
| s1mb generalization diverse choice |
Choice |
94.7 |
100.0 |
70.3 |
|
来源 ↗ |
| openbookqa |
Choice |
94.4 |
97.2 |
54.7 |
|
来源 ↗ |
| laya / enron spam |
Noul |
94.0 |
100.0 |
55.5 |
|
来源 ↗ |
| s1mb generalization contextual noul |
Noul |
94.0 |
98.0 |
63.2 |
|
来源 ↗ |
| open jev / vizdoom basic v1 |
Score |
93.3 |
100.0 |
32.3 |
|
来源 ↗ |
| miqa |
Choice |
93.2 |
95.5 |
63.0 |
|
来源 ↗ |
| s1mb generalization diverse noul |
Noul |
92.0 |
100.0 |
60.5 |
|
来源 ↗ |
| open jev / painting geometry v1 |
Choice |
89.5 |
100.0 |
48.2 |
|
来源 ↗ |
| snli |
Choice |
88.7 |
98.4 |
47.8 |
|
来源 ↗ |
| massive |
Choice |
87.9 |
96.7 |
65.7 |
|
来源 ↗ |
| mtop |
Choice |
87.4 |
87.4 |
59.7 |
|
来源 ↗ |
| banking77 |
Choice |
87.4 |
93.7 |
55.8 |
|
来源 ↗ |
| bbq |
Choice |
86.7 |
98.3 |
54.5 |
|
来源 ↗ |
| open jev / context retention control v1 |
Noul |
86.0 |
100.0 |
21.2 |
|
来源 ↗ |
| scitail |
Noul |
84.0 |
98.0 |
54.3 |
|
来源 ↗ |
| sdoh nli |
Noul |
84.0 |
92.0 |
61.6 |
|
来源 ↗ |
| hwu64 |
Choice |
83.0 |
92.5 |
58.1 |
|
来源 ↗ |
| open jev / customer control v1 |
Choice |
82.6 |
82.6 |
65.1 |
|
来源 ↗ |
| open jev / workflow controls v1 / customer service |
Noul |
82.2 |
100.0 |
43.7 |
|
来源 ↗ |
| impli |
Noul |
82.0 |
88.0 |
47.7 |
|
来源 ↗ |
| open jev / painting geometry v1 |
Score |
81.0 |
100.0 |
17.2 |
|
来源 ↗ |
| laya / ag news |
Choice |
80.6 |
94.0 |
75.3 |
|
来源 ↗ |
| open jev / workflow controls v1 / agent trace observability |
Noul |
78.7 |
100.0 |
29.6 |
|
来源 ↗ |
| open jev / ir control v1 |
Score |
78.4 |
98.6 |
12.8 |
|
来源 ↗ |
| contract nli |
Choice |
78.3 |
80.3 |
33.7 |
|
来源 ↗ |
| open jev / amount extraction control v1 |
Noul |
78.1 |
100.0 |
35.2 |
|
来源 ↗ |
| cladder |
Noul |
78.0 |
94.0 |
19.4 |
|
来源 ↗ |
| followir robust04 |
Noul |
77.3 |
90.9 |
47.7 |
|
来源 ↗ |
| gretel pii |
Choice |
77.3 |
94.5 |
44.3 |
|
来源 ↗ |
| lonli |
Choice |
77.0 |
86.9 |
40.5 |
|
来源 ↗ |
| open jev / amount extraction control v1 |
Choice |
76.5 |
100.0 |
56.2 |
|
来源 ↗ |
| quartz |
Choice |
76.2 |
88.1 |
50.2 |
|
来源 ↗ |
| ethos |
Noul |
76.0 |
76.0 |
43.7 |
|
来源 ↗ |
| hans |
Noul |
76.0 |
100.0 |
44.8 |
|
来源 ↗ |
| open jev / ir control v1 |
Choice |
76.0 |
88.0 |
45.8 |
|
来源 ↗ |
| s1mb generalization diverse score |
Score |
75.1 |
95.3 |
43.8 |
|
来源 ↗ |
| arct |
Choice |
74.5 |
89.4 |
40.5 |
|
来源 ↗ |
| hatecheck |
Noul |
74.0 |
98.0 |
47.9 |
|
来源 ↗ |
| paws |
Noul |
74.0 |
88.0 |
46.9 |
|
来源 ↗ |
| ruletaker |
Noul |
74.0 |
96.0 |
29.2 |
|
来源 ↗ |
| s1mb generalization contextual score |
Score |
73.9 |
92.7 |
43.1 |
|
来源 ↗ |
| argument quality |
Choice |
73.5 |
95.9 |
39.4 |
|
来源 ↗ |
| open jev / workflow controls v1 / invoice processing |
Noul |
71.1 |
100.0 |
36.1 |
|
来源 ↗ |
| open jev / painting geometry v1 |
Noul |
70.5 |
100.0 |
29.2 |
|
来源 ↗ |
| open jev / reasoning control v1 |
Choice |
70.0 |
95.0 |
29.8 |
|
来源 ↗ |
| winobias |
Choice |
69.6 |
93.5 |
30.9 |
|
来源 ↗ |
| few nerd |
Choice |
69.2 |
86.4 |
45.1 |
|
来源 ↗ |
| qasper |
Noul |
68.0 |
88.0 |
42.5 |
|
来源 ↗ |
| scone |
Choice |
68.0 |
92.0 |
36.1 |
|
来源 ↗ |
| laya / typed decisions |
Score |
66.7 |
67.0 |
28.0 |
|
来源 ↗ |
| ethics |
Choice |
66.7 |
83.3 |
21.7 |
|
来源 ↗ |
| canttalk |
Noul |
66.0 |
86.0 |
45.0 |
|
来源 ↗ |
| open jev / reasoning control v1 |
Score |
65.4 |
90.7 |
28.7 |
|
来源 ↗ |
| gsm8k |
Choice |
64.3 |
81.4 |
25.2 |
|
来源 ↗ |
| robust lr |
Choice |
61.7 |
66.7 |
19.7 |
|
来源 ↗ |
| open jev / workflow controls v1 / security incidents |
Noul |
59.8 |
100.0 |
38.5 |
|
来源 ↗ |
| scicite |
Choice |
59.7 |
71.0 |
44.6 |
|
来源 ↗ |
| creak |
Noul |
58.0 |
88.0 |
39.1 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanomsmarco |
Noul |
57.4 |
62.5 |
38.3 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonq |
Noul |
57.3 |
69.7 |
40.4 |
|
来源 ↗ |
| go emotions |
Noul |
56.6 |
61.3 |
38.2 |
|
来源 ↗ |
| defeasible nli |
Choice |
56.5 |
80.4 |
34.1 |
|
来源 ↗ |
| crows pairs |
Choice |
56.3 |
70.8 |
31.1 |
|
来源 ↗ |
| babi nli |
Noul |
56.0 |
84.0 |
38.3 |
|
来源 ↗ |
| open jev / customer control v1 |
Score |
55.5 |
99.3 |
37.9 |
|
来源 ↗ |
| aegis2 |
Noul |
54.8 |
62.3 |
33.6 |
|
来源 ↗ |
| open jev / entity alignment control v1 |
Score |
49.4 |
100.0 |
9.7 |
|
来源 ↗ |
| synthetic relevance / nanocoir |
Noul |
48.4 |
73.0 |
34.4 |
|
来源 ↗ |
| open jev / snake v1 |
Choice |
46.4 |
83.5 |
9.2 |
|
来源 ↗ |
| laya / typed decisions |
Noul |
45.8 |
47.8 |
25.1 |
|
来源 ↗ |
| boardgameqa |
Choice |
45.3 |
62.5 |
16.6 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanomsmarco |
Score |
43.8 |
52.6 |
19.2 |
|
来源 ↗ |
| esci |
Choice |
42.5 |
46.6 |
13.8 |
|
来源 ↗ |
| lexcomp |
Noul |
42.0 |
52.0 |
23.3 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanotouche2020 |
Noul |
41.5 |
47.0 |
27.3 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanofiqa2018 |
Noul |
41.4 |
54.0 |
29.2 |
|
来源 ↗ |
| aqua rat |
Choice |
40.9 |
77.5 |
11.5 |
|
来源 ↗ |
| open jev / reasoning control v1 |
Noul |
40.4 |
100.0 |
25.7 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanofiqa2018 |
Score |
40.1 |
46.7 |
16.2 |
|
来源 ↗ |
| stepgame |
Choice |
39.3 |
82.1 |
18.6 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanodbpedia |
Noul |
38.9 |
52.1 |
29.5 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanotouche2020 |
Score |
38.5 |
44.3 |
21.4 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanodbpedia |
Score |
38.1 |
45.0 |
18.7 |
|
来源 ↗ |
| wiqa |
Choice |
36.5 |
58.7 |
20.4 |
|
来源 ↗ |
| patent similarity |
Score |
35.4 |
48.3 |
18.7 |
|
来源 ↗ |
| ethics |
Noul |
34.2 |
71.0 |
24.8 |
|
来源 ↗ |
| open jev / silent failure control v1 |
Noul |
34.0 |
100.0 |
40.5 |
|
来源 ↗ |
| civil comments |
Noul |
33.6 |
36.7 |
21.6 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonfcorpus |
Noul |
33.4 |
46.2 |
25.5 |
|
来源 ↗ |
| laya / typed decisions |
Choice |
32.4 |
34.5 |
19.5 |
|
来源 ↗ |
| followir core17 |
Noul |
32.4 |
46.6 |
21.7 |
|
来源 ↗ |
| logical entailment |
Noul |
32.0 |
72.0 |
12.4 |
|
来源 ↗ |
| winowhy |
Noul |
32.0 |
62.0 |
19.4 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanohotpotqa |
Score |
31.6 |
63.6 |
14.8 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonq |
Score |
28.6 |
49.4 |
10.2 |
|
来源 ↗ |
| temporal nli |
Choice |
27.4 |
80.7 |
19.5 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonfcorpus |
Score |
27.4 |
38.6 |
11.6 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoarguana |
Noul |
26.7 |
53.5 |
16.4 |
|
来源 ↗ |
| ud ewt |
Choice |
26.4 |
56.7 |
13.9 |
|
来源 ↗ |
| open jev / drone control v1 |
Choice |
25.4 |
100.0 |
5.6 |
|
来源 ↗ |
| open jev / drone control v1 |
Score |
25.2 |
99.2 |
2.0 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoscidocs |
Noul |
23.9 |
41.7 |
17.6 |
|
来源 ↗ |
| sgd |
Choice |
23.7 |
96.6 |
25.2 |
|
来源 ↗ |
| open jev / snake v1 |
Noul |
21.6 |
100.0 |
9.9 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanohotpotqa |
Noul |
21.5 |
70.8 |
28.4 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoscidocs |
Score |
20.5 |
36.7 |
7.7 |
|
来源 ↗ |
| fol nli |
Choice |
18.0 |
54.1 |
10.9 |
|
来源 ↗ |
| spartqa |
Choice |
17.7 |
38.7 |
4.9 |
|
来源 ↗ |
| tracie |
Noul |
16.0 |
34.0 |
9.1 |
|
来源 ↗ |
| argument quality |
Noul |
12.3 |
23.9 |
5.5 |
|
来源 ↗ |
| plane |
Noul |
12.0 |
100.0 |
12.0 |
|
来源 ↗ |
| corr2cause |
Choice |
11.7 |
91.7 |
4.1 |
|
来源 ↗ |
| nlsat |
Noul |
10.0 |
22.0 |
2.3 |
|
来源 ↗ |
| followir news21 |
Noul |
9.3 |
42.7 |
15.3 |
|
来源 ↗ |
| open jev / tic tac toe v1 |
Choice |
6.8 |
48.5 |
3.5 |
|
来源 ↗ |
| synthetic relevance / nanocoir |
Score |
5.2 |
44.3 |
4.6 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoarguana |
Score |
2.3 |
39.3 |
8.0 |
|
来源 ↗ |
| hh rlhf |
Choice |
0.0 |
23.1 |
3.1 |
|
来源 ↗ |
| poem sentiment |
Choice |
0.0 |
57.1 |
4.7 |
|
来源 ↗ |