Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

C3

Benchmarks

Task NameDataset NameSOTA ResultTrend
Reading ComprehensionC3
Accuracy93.5
89
UnderstandingC3
Score68.94
20
Reading ComprehensionC3 (test)
Accuracy77.38
16
Chinese Reading ComprehensionC3
Score57.97
10
Chinese ReasoningC3
Accuracy96.88
8
Multi-choice Question AnsweringC3
Accuracy44.6
8
Spoken Dialogue EvaluationC3 ZH
Phonetic Error9.19
7
Spoken Dialogue EvaluationC3 EN
Phonetic Score48.28
7
DirectnessC3 news community level post-cutoff (test)
Win Percentage61
6
Camera pose estimationC3
Angular Recall @ 5°65.91
5
Correspondence EstimationC3 clean
PCK @ 1%20.42
5
Reading ComprehensionC3
EM83.1
3
Showing 12 of 12 rows