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

Mathematical Reasoning on MGSM

93.2Accuracy

DeepSeek V3.2

46.97258.973570.97582.9765May 17, 2023Nov 16, 2023May 17, 2024Nov 17, 2024May 19, 2025Nov 18, 2025May 21, 2026
Updated 9d ago

Evaluation Results

MethodLinks
2026.05
93.2--
2026.02
91.7--
2024.07
91.6--
2024.07
91.6--
2026.02
91.6--
2026.05
90.9-5.9
2026.05
90.8-17.9
2024.07
90.5--
2026.02
90.5--
2026.05
90.4-5.3
2026.02
90--
2026.02
89.6--
2025.02
89.3--
89.01--
2026.05
88.9-3.6
2026.02
88.4--
2025.02
88.16--
2026.05
88-14.3
2026.05
87.8-2.3
2026.02
87.4--
87.36--
2026.05
87.2-1.6
2024.07
86.9--
2026.05
86.4-0.7
2026.05
86.1-0.4
2026.02
86--
2024.07
85.9--
2026.05
85.8--
2026.05
83.1-4.3
2026.05
82.5-7.1
2026.05
82.4--
2025.02
82.27--
2026.05
82.1-3.3
2026.05
81.5-2.5
2026.05
81.4-5.5
2026.05
81.2-2.1
2026.05
80.8-1.6
2026.05
80.2-0.1
2026.05
79.5--
2026.05
79.5-0
2026.05
78.5-1.9
2025.02
78.37--
2026.05
78.3-1.6
2026.05
77.64--
2026.05
77.5-0.6
2026.05
77--
2026.02
76.6--
2026.05
76.56--
2026.02
76.1--
2023.05
75.9--
2026.02
73--
2026.02
72.7--
2026.05
72.4--
2023.05
72.2--
2026.02
71.9--
2024.07
71.1--
2026.05
70.96--
2026.05
70.12--
2024.07
68.9--
2026.02
67.3--
2025.02
66.9--
2025.07
65.71--
2025.07
64.13--
2025.07
63.92--
2025.07
61.87--
2025.02
60.8--
2026.02
60.5--
2023.05
60.4--
2025.02
59--
2026.02
58.9--
2025.07
58.45--
2026.05
58.45--
2026.01
58.44--
2025.07
58.03--
2026.05
58.03--
2025.07
57.25--
2025.07
56.29--
2026.05
56.29--
2025.07
55.28--
2026.05
55.28--
2025.07
55.09--
2025.07
53.89--
2025.07
53.44--
2024.07
53.2--
2026.01
53.2--
2026.05
52.36--
2026.01
52--
2025.07
51.95--
2024.07
51.4--
2025.07
50.88--
2025.07
50.75--
2025.07
50.43--
2025.07
50.29--
2025.07
50.03--
2025.07
50.03--
2023.05
49.9--
2025.07
49.63--
2025.07
49.6--
2025.07
48.99--
2025.07
48.75--
Showing 100 of 240 rows