Arithmetic Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
0.971Accuracy
18
Jun 26, 2026
98.5Accuracy
16
Jun 30, 2026
84.3Accuracy
16
Jun 30, 2026
83.1Accuracy
16
Jun 30, 2026
62.1Pass@1
15
Feb 26, 2026
71.1Pass@1
15
Feb 26, 2026
83.3ACC
15
Jun 2, 2026
91.6Accuracy
14
Jun 26, 2026
87.64Pass@1
14
Feb 26, 2026
64.8Accuracy
14
Feb 26, 2026
76.27Validation Reward (mean@4)
13
Jun 23, 2026
69.3Accuracy
12
Feb 26, 2026
85.3Performance
11
Mar 24, 2026
37.1Accuracy
10
Feb 26, 2026
90.25Accuracy
10
Feb 26, 2026
94.3Accuracy
10
Feb 27, 2026
100Success Rate
9
Feb 26, 2026
97.3Accuracy
9
Feb 26, 2026
93.2Avg@16
8
Jun 4, 2026
70.66Accuracy
8
Mar 18, 2026
71.5Pass@1 (Greedy)
7
Apr 16, 2026
43.4Average Accuracy
7
Mar 4, 2026
90.3Accuracy
7
Feb 26, 2026
84.33Addition Accuracy
6
Jun 9, 2026
92Pass@1
6
Apr 15, 2026