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

Mathematical Reasoning on GSM8K OOD (test)

94.98Accuracy

MultiRole-R1

76.62481.389586.15590.9205Jul 27, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.07
94.9886.93
2025.07
93.5881.93
2025.07
93.4387.24
2025.07
91.9785.98
2025.07
91.8786.36
2025.07
91.8465.32
2025.07
91.6270.23
2025.07
91.6191.62
2025.07
91.576.6
2025.07
90.9268.69
2025.07
90.0275.07
2025.07
89.7988.45
2025.07
88.7680.94
2025.07
87.4985.31
2025.07
87.3682.46
2025.07
87.0180.37
2025.07
86.3481.81
2025.07
86.0281.53
2025.07
85.6672.87
2025.07
85.5882.16
2025.07
85.5870.68
2025.07
85.4170.49
2025.07
84.9182.31
2025.07
84.7380.24
2025.07
83.384.12
2025.07
82.0581.79
2025.07
81.5167.56
2025.07
80.9581.24
2025.07
80.6274.87
2025.07
80.4868.08
2025.07
78.8776.52
2025.07
77.3375.02