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

Arithmetic Reasoning on SVAMP (Accuracy)

96.01Accuracy

Graph-GRPO

9.263631.784354.30576.8257Mar 21, 2022Nov 16, 2022Jul 15, 2023Mar 11, 2024Nov 7, 2024Jul 5, 2025Mar 3, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
96.01
94.7
2024.04
94.2
2026.03
93.1
2024.04
92.6
91.04
2026.03
90.58
2024.04
90.4
2024.04
90.3
2026.03
90.11
2026.03
88.91
2026.03
88.59
2024.05
87.6
2026.03
87.54
2026.03
87.36
2026.03
87.17
2022.03
86.8
2022.03
86.6
2026.03
86.55
2024.05
86.2
2023.06
86
2023.06
85.6
2023.06
85
2026.03
84.01
2024.05
83.6
2024.05
83.1
2024.04
82.9
2024.05
82.8
2024.05
82.6
2024.04
82.5
2023.06
81.4
2024.04
80.9
2024.04
80.7
2023.06
79.5
2024.04
79.3
2022.03
79
2024.04
78.5
2023.06
77.7
2022.03
75.8
2023.06
69.9
2022.10
69.5
2023.06
69.5
2022.10
68.9
2023.06
68.9
2024.05
67.2
2022.10
65.7
2023.06
65.7
2022.10
63.7
2023.06
62.1
2022.10
58.8
2023.06
58.8
57.4
57.4
2024.05
57.1
2024.05
54.9
2022.03
54.5
2022.03
53.3
2022.03
39.8
2022.03
38.9
2022.03
19.4
2022.03
12.6