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

MGSM

Benchmarks

Task NameDataset NameSOTA ResultTrend
Mathematical ReasoningMGSM
Accuracy93.2
245
Multilingual Mathematical ReasoningMGSM (test)
Accuracy93.4
109
Mathematical ReasoningMGSM (test)
Accuracy (ZH)88
80
Mathematical ReasoningMGSM
Accuracy (Bn)88.8
66
Multilingual Mathematical ReasoningMGSM
Accuracy94.9
64
Question AnsweringMGSM
AUC90.79
51
Hallucination DetectionMGSM
AUC84.62
42
Mathematical ReasoningMGSM Rev2
Overall Result Accuracy99.4
36
Multilingual Mathematical ReasoningMGSM 1.0 (test)
Accuracy (ru)69.6
35
ReasoningMGSM
Accuracy94.4
29
Multilingual Mathematical ReasoningMGSM Thai (test)
Accuracy41.6
25
Multilingual Math ReasoningMGSM
Mean@384.43
23
Math reasoningmGSM v2
Accuracy (Seen)78
21
Multilingual mathematical reasoningMGSM
Speedup (de)6.08
20
Mathematical ReasoningMGSM non-EU languages (test)
Accuracy91.4
16
Mathematical ReasoningMGSM 24 official EU languages
Accuracy93
14
Mathematical ReasoningMGSM Bangla
Accuracy (Original)0.88
13
Multilingual Mathematical ReasoningMGSM Rev2 (all)
Accuracy85.9
12
Multilingual Mathematical ReasoningMGSM out-of-domain Rev2
Accuracy83.6
12
Multilingual Mathematical ReasoningMGSM in-domain Rev2
Accuracy90.4
12
Mathematical ReasoningMGSM
Average@1044
12
Mathematical ReasoningMGSM ALL 1.0 (test)
Accuracy75.93
12
Mathematical ReasoningMGSM EN 1.0 (test)
Accuracy94
12
Machine TranslationMGSM
Bn Score72.89
12
Mathematical ReasoningBn-MGSM (test)
Accuracy89.2
12
Showing 25 of 56 rows