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

ARC

Benchmarks

Task NameDataset NameSOTA ResultTrend
Question AnsweringARC Challenge
Accuracy96.3
906
Question AnsweringARC Easy
Accuracy98.2
597
Question AnsweringARC-E
Accuracy95.23
416
Question AnsweringARC Easy
Normalized Acc96.4
389
Science Question AnsweringARC Challenge
Accuracy96
342
Question AnsweringARC
Accuracy94.6
230
Science Question AnsweringARC-C
Accuracy96.3
193
Question AnsweringARC-C
Accuracy94.1
192
Commonsense ReasoningARC Challenge
Accuracy93.8
190
Multiple Choice Question AnsweringARC Easy
Accuracy99.7
188
ReasoningARC Easy
Accuracy96.63
187
Science Question AnsweringARC-E
Accuracy97.53
184
Commonsense ReasoningARC-C
Accuracy96.3
172
Science Question AnsweringARC Easy
Accuracy98
155
Question AnsweringARC Challenge
Accuracy (ARC)87.3
142
Multiple-choice Question AnsweringARC Challenge
Acc74.7
118
Commonsense ReasoningARC-E
Accuracy96.4
106
Scientific ReasoningARC Challenge
Accuracy92.5
94
ReasoningARC
Accuracy92.34
94
ReasoningARC Challenge
Accuracy97.2
93
Question AnsweringARC-C
Accuracy0.71
87
Question AnsweringARC Challenge
Normalized Accuracy59
86
ReasoningARC-c
Accuracy90.36
80
Question AnsweringARC Challenge (val)
Accuracy93.3
76
Question AnsweringARC Challenge (test)
Accuracy91.2
73
Showing 25 of 272 rows
...