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

Multilingual Mathematical Reasoning on MGSM out-of-domain Rev2

83.6Accuracy

SFT

51.56859.88468.276.516Jun 13, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
83.688.91.8
2026.06
83.690.51.4
2026.06
82.990.31.6
2026.06
80.43617.6
2026.06
77.434.517.8
2026.06
65.564.82.3
2026.06
6517.721.1
2026.06
64.21923.3
2026.06
63.661.62.6
2026.06
58.362.84.2
2026.06
57.858.634.7
2026.06
52.860.318.4