| hans |
Noul |
100.0 |
100.0 |
47.8 |
|
来源 ↗ |
| open jev / amount extraction control v1 |
Choice |
100.0 |
100.0 |
59.1 |
|
来源 ↗ |
| open jev / browser control v1 |
Choice |
100.0 |
100.0 |
72.5 |
|
来源 ↗ |
| open jev / citation control v1 |
Choice |
100.0 |
100.0 |
49.2 |
|
来源 ↗ |
| open jev / context retention control v1 |
Noul |
100.0 |
100.0 |
27.7 |
|
来源 ↗ |
| open jev / customer control v1 |
Noul |
100.0 |
100.0 |
62.8 |
|
来源 ↗ |
| open jev / drone control v1 |
Noul |
100.0 |
100.0 |
59.1 |
|
来源 ↗ |
| open jev / email selection control v1 |
Choice |
100.0 |
100.0 |
31.4 |
|
来源 ↗ |
| open jev / email selection control v1 |
Noul |
100.0 |
100.0 |
49.2 |
|
来源 ↗ |
| open jev / entity alignment control v1 |
Noul |
100.0 |
100.0 |
44.5 |
|
来源 ↗ |
| open jev / ir control v1 |
Noul |
100.0 |
100.0 |
54.8 |
|
来源 ↗ |
| open jev / mailroom control v1 |
Choice |
100.0 |
100.0 |
86.5 |
|
来源 ↗ |
| open jev / painting geometry v1 |
Choice |
100.0 |
100.0 |
53.2 |
|
来源 ↗ |
| open jev / painting geometry v1 |
Noul |
100.0 |
100.0 |
35.4 |
|
来源 ↗ |
| open jev / phone extraction control v1 |
Choice |
100.0 |
100.0 |
39.4 |
|
来源 ↗ |
| open jev / phone extraction control v1 |
Noul |
100.0 |
100.0 |
57.4 |
|
来源 ↗ |
| open jev / silent failure control v1 |
Noul |
100.0 |
100.0 |
43.8 |
|
来源 ↗ |
| open jev / sponsor segment control v1 |
Choice |
100.0 |
100.0 |
60.7 |
|
来源 ↗ |
| open jev / vizdoom basic v1 |
Noul |
100.0 |
100.0 |
46.5 |
|
来源 ↗ |
| open jev / workflow controls v1 / agent trace observability |
Noul |
100.0 |
100.0 |
34.1 |
|
来源 ↗ |
| s1mb generalization contextual choice |
Choice |
100.0 |
100.0 |
86.0 |
|
来源 ↗ |
| s1mb generalization diverse choice |
Choice |
100.0 |
100.0 |
74.3 |
|
来源 ↗ |
| s1mb generalization diverse noul |
Noul |
100.0 |
100.0 |
66.2 |
|
来源 ↗ |
| open jev / vizdoom basic v1 |
Score |
100.0 |
100.0 |
35.8 |
|
来源 ↗ |
| open jev / entity alignment control v1 |
Score |
100.0 |
100.0 |
12.9 |
|
来源 ↗ |
| open jev / painting geometry v1 |
Score |
99.9 |
100.0 |
22.2 |
|
来源 ↗ |
| open jev / ir control v1 |
Score |
99.7 |
99.7 |
16.5 |
|
来源 ↗ |
| arc |
Choice |
98.5 |
98.5 |
69.7 |
|
来源 ↗ |
| open jev / customer control v1 |
Score |
98.4 |
99.3 |
40.5 |
|
来源 ↗ |
| s1mb generalization contextual noul |
Noul |
98.0 |
100.0 |
68.3 |
|
来源 ↗ |
| scitail |
Noul |
98.0 |
98.0 |
57.5 |
|
来源 ↗ |
| miqa |
Choice |
97.7 |
97.7 |
67.3 |
|
来源 ↗ |
| open jev / snake v1 |
Noul |
97.3 |
100.0 |
14.3 |
|
来源 ↗ |
| open jev / workflow controls v1 / customer service |
Noul |
97.3 |
100.0 |
49.2 |
|
来源 ↗ |
| clinc |
Choice |
97.2 |
99.3 |
65.8 |
|
来源 ↗ |
| open jev / amount extraction control v1 |
Noul |
96.9 |
100.0 |
38.6 |
|
来源 ↗ |
| dbpedia |
Choice |
96.6 |
98.9 |
86.0 |
|
来源 ↗ |
| openbookqa |
Choice |
95.8 |
97.2 |
59.2 |
|
来源 ↗ |
| open jev / workflow controls v1 / invoice processing |
Noul |
95.8 |
100.0 |
41.4 |
|
来源 ↗ |
| babi nli |
Noul |
94.0 |
94.0 |
42.8 |
|
来源 ↗ |
| laya / enron spam |
Noul |
94.0 |
100.0 |
60.8 |
|
来源 ↗ |
| scone |
Choice |
94.0 |
98.0 |
41.5 |
|
来源 ↗ |
| winobias |
Choice |
93.5 |
93.5 |
36.5 |
|
来源 ↗ |
| bbq |
Choice |
93.3 |
98.3 |
60.5 |
|
来源 ↗ |
| open jev / mailroom control v1 |
Noul |
93.3 |
100.0 |
68.3 |
|
来源 ↗ |
| s1mb generalization diverse score |
Score |
93.3 |
96.3 |
47.4 |
|
来源 ↗ |
| open jev / reasoning control v1 |
Score |
92.5 |
95.1 |
32.5 |
|
来源 ↗ |
| open jev / drone control v1 |
Score |
92.3 |
99.2 |
4.4 |
|
来源 ↗ |
| plane |
Noul |
92.0 |
100.0 |
11.8 |
|
来源 ↗ |
| banking77 |
Choice |
91.6 |
93.7 |
60.3 |
|
来源 ↗ |
| quartz |
Choice |
90.5 |
90.5 |
54.8 |
|
来源 ↗ |
| s1mb generalization contextual score |
Score |
90.5 |
97.3 |
47.3 |
|
来源 ↗ |
| open jev / workflow controls v1 / security incidents |
Noul |
90.3 |
100.0 |
43.8 |
|
来源 ↗ |
| lonli |
Choice |
90.2 |
96.7 |
46.3 |
|
来源 ↗ |
| open jev / reasoning control v1 |
Choice |
90.0 |
100.0 |
35.4 |
|
来源 ↗ |
| sdoh nli |
Noul |
90.0 |
94.0 |
65.0 |
|
来源 ↗ |
| massive |
Choice |
89.6 |
96.7 |
69.6 |
|
来源 ↗ |
| creak |
Noul |
88.0 |
90.0 |
43.4 |
|
来源 ↗ |
| open jev / ir control v1 |
Choice |
88.0 |
88.0 |
51.3 |
|
来源 ↗ |
| qasper |
Noul |
88.0 |
88.0 |
47.4 |
|
来源 ↗ |
| ruletaker |
Noul |
88.0 |
100.0 |
32.5 |
|
来源 ↗ |
| mtop |
Choice |
86.3 |
88.6 |
63.7 |
|
来源 ↗ |
| argument quality |
Choice |
85.7 |
95.9 |
45.5 |
|
来源 ↗ |
| hwu64 |
Choice |
85.1 |
93.6 |
63.2 |
|
来源 ↗ |
| impli |
Noul |
84.0 |
88.0 |
52.0 |
|
来源 ↗ |
| laya / ag news |
Choice |
83.6 |
94.0 |
75.4 |
|
来源 ↗ |
| open jev / customer control v1 |
Choice |
82.6 |
82.6 |
68.6 |
|
来源 ↗ |
| open jev / reasoning control v1 |
Noul |
82.0 |
100.0 |
29.6 |
|
来源 ↗ |
| gsm8k |
Choice |
81.4 |
95.7 |
30.8 |
|
来源 ↗ |
| arct |
Choice |
80.8 |
89.4 |
43.7 |
|
来源 ↗ |
| logical entailment |
Noul |
80.0 |
80.0 |
14.1 |
|
来源 ↗ |
| contract nli |
Choice |
79.9 |
80.3 |
37.8 |
|
来源 ↗ |
| snli |
Choice |
79.0 |
98.4 |
53.1 |
|
来源 ↗ |
| defeasible nli |
Choice |
78.3 |
87.0 |
37.9 |
|
来源 ↗ |
| hatecheck |
Noul |
78.0 |
98.0 |
52.2 |
|
来源 ↗ |
| open jev / snake v1 |
Choice |
77.3 |
83.5 |
13.4 |
|
来源 ↗ |
| few nerd |
Choice |
77.1 |
86.4 |
48.9 |
|
来源 ↗ |
| paws |
Noul |
74.0 |
90.0 |
48.8 |
|
来源 ↗ |
| crows pairs |
Choice |
70.8 |
70.8 |
34.1 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonq |
Noul |
68.7 |
71.2 |
44.4 |
|
来源 ↗ |
| ethics |
Choice |
66.7 |
83.3 |
24.6 |
|
来源 ↗ |
| canttalk |
Noul |
66.0 |
92.0 |
48.1 |
|
来源 ↗ |
| ethics |
Noul |
65.7 |
75.6 |
28.6 |
|
来源 ↗ |
| synthetic relevance / nanocoir |
Noul |
65.6 |
73.0 |
38.6 |
|
来源 ↗ |
| gretel pii |
Choice |
65.3 |
94.5 |
48.0 |
|
来源 ↗ |
| followir robust04 |
Noul |
62.9 |
90.9 |
51.2 |
|
来源 ↗ |
| cladder |
Noul |
62.0 |
94.0 |
22.8 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanomsmarco |
Noul |
62.0 |
63.0 |
41.7 |
|
来源 ↗ |
| ethos |
Noul |
60.0 |
76.0 |
47.1 |
|
来源 ↗ |
| robust lr |
Choice |
60.0 |
66.7 |
23.4 |
|
来源 ↗ |
| tracie |
Noul |
60.0 |
60.0 |
10.4 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanohotpotqa |
Noul |
58.5 |
77.7 |
32.2 |
|
来源 ↗ |
| aegis2 |
Noul |
58.2 |
64.9 |
36.6 |
|
来源 ↗ |
| laya / typed decisions |
Score |
55.0 |
76.0 |
32.2 |
|
来源 ↗ |
| scicite |
Choice |
54.8 |
71.0 |
46.0 |
|
来源 ↗ |
| aqua rat |
Choice |
53.5 |
83.1 |
14.8 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanohotpotqa |
Score |
53.5 |
63.6 |
16.4 |
|
来源 ↗ |
| lexcomp |
Noul |
52.0 |
60.0 |
24.9 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanomsmarco |
Score |
50.7 |
54.9 |
21.3 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanofiqa2018 |
Noul |
50.2 |
58.0 |
32.3 |
|
来源 ↗ |
| boardgameqa |
Choice |
50.0 |
75.0 |
19.8 |
|
来源 ↗ |
| temporal nli |
Choice |
50.0 |
83.9 |
21.4 |
|
来源 ↗ |
| sgd |
Choice |
49.1 |
96.6 |
26.4 |
|
来源 ↗ |
| winowhy |
Noul |
48.0 |
64.0 |
21.7 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoarguana |
Noul |
47.9 |
53.5 |
19.2 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanodbpedia |
Score |
47.7 |
47.7 |
20.4 |
|
来源 ↗ |
| laya / typed decisions |
Noul |
46.8 |
49.2 |
27.5 |
|
来源 ↗ |
| go emotions |
Noul |
46.3 |
63.4 |
40.4 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanotouche2020 |
Noul |
46.0 |
47.0 |
29.5 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanodbpedia |
Noul |
45.6 |
52.1 |
32.3 |
|
来源 ↗ |
| stepgame |
Choice |
45.2 |
82.1 |
20.9 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonq |
Score |
44.7 |
49.4 |
11.1 |
|
来源 ↗ |
| ud ewt |
Choice |
43.3 |
64.1 |
15.6 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoscidocs |
Noul |
43.0 |
46.6 |
20.4 |
|
来源 ↗ |
| civil comments |
Noul |
42.8 |
42.8 |
23.2 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanofiqa2018 |
Score |
42.4 |
46.7 |
17.8 |
|
来源 ↗ |
| patent similarity |
Score |
40.9 |
48.3 |
20.6 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanotouche2020 |
Score |
38.6 |
47.5 |
23.2 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonfcorpus |
Noul |
38.1 |
49.7 |
28.1 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoarguana |
Score |
38.0 |
43.1 |
9.0 |
|
来源 ↗ |
| wiqa |
Choice |
36.5 |
60.3 |
21.2 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanonfcorpus |
Score |
36.1 |
38.6 |
13.1 |
|
来源 ↗ |
| laya / typed decisions |
Choice |
35.5 |
35.5 |
21.5 |
|
来源 ↗ |
| followir core17 |
Noul |
35.0 |
46.6 |
23.6 |
|
来源 ↗ |
| spartqa |
Choice |
32.3 |
91.9 |
7.4 |
|
来源 ↗ |
| esci |
Choice |
31.5 |
46.6 |
15.0 |
|
来源 ↗ |
| synthetic relevance / nanocoir |
Score |
31.1 |
44.3 |
5.3 |
|
来源 ↗ |
| followir news21 |
Noul |
28.7 |
48.0 |
17.4 |
|
来源 ↗ |
| fol nli |
Choice |
27.9 |
68.8 |
12.9 |
|
来源 ↗ |
| synthetic relevance / nanobeir / nanoscidocs |
Score |
25.3 |
36.7 |
8.5 |
|
来源 ↗ |
| open jev / drone control v1 |
Choice |
21.1 |
100.0 |
9.3 |
|
来源 ↗ |
| argument quality |
Noul |
18.4 |
26.9 |
6.6 |
|
来源 ↗ |
| poem sentiment |
Choice |
14.3 |
68.6 |
5.3 |
|
来源 ↗ |
| corr2cause |
Choice |
10.0 |
91.7 |
4.3 |
|
来源 ↗ |
| open jev / tic tac toe v1 |
Choice |
9.1 |
54.5 |
4.4 |
|
来源 ↗ |
| nlsat |
Noul |
8.0 |
22.0 |
2.8 |
|
来源 ↗ |
| hh rlhf |
Choice |
7.7 |
33.3 |
3.4 |
|
来源 ↗ |