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 Challenge
Accuracy (ARC)87.3
598
Question AnsweringARC Easy
Accuracy98.2
597
Question AnsweringARC-E
Accuracy95.23
523
Question AnsweringARC Easy
Normalized Acc96.4
391
Science Question AnsweringARC Challenge
Accuracy96
354
Science Question AnsweringARC-C
Accuracy96.3
261
Question AnsweringARC-C
Accuracy94.1
258
Multiple Choice Question AnsweringARC Easy
Accuracy99.7
257
ReasoningARC
Accuracy94.5
245
Commonsense ReasoningARC Challenge
Accuracy93.8
243
Science Question AnsweringARC-E
Accuracy97.53
240
ReasoningARC Easy
Accuracy96.63
233
Question AnsweringARC
Accuracy94.6
230
Commonsense ReasoningARC-C
Accuracy96.3
215
Question AnsweringARC Easy
Accuracy90.48
210
Science Question AnsweringARC Easy
Accuracy98
162
Question AnsweringARC (test)
Accuracy90.5
153
Commonsense ReasoningARC-E
Accuracy96.4
152
Multiple-choice Question AnsweringARC Challenge
Acc74.7
133
Question AnsweringARC-C
Accuracy0.71
116
Scientific ReasoningARC Challenge
Accuracy92.5
115
ReasoningARC-c
Accuracy90.85
112
Science Question AnsweringARC Challenge
Accuracy93
108
Question AnsweringARC Challenge
Normalized Accuracy59
105
Showing 25 of 418 rows
...