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

Math Reasoning on OlympiadBench (Accuracy)

90.7Accuracy

APRM

-0.19623.4024770.598Nov 28, 2025Dec 9, 2025Dec 20, 2025Jan 1, 2026Jan 12, 2026Jan 23, 2026Feb 4, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2025.11
90.7
2025.11
90.4
2025.11
89.4
2025.11
85.6
2025.11
85.2
2025.11
84.7
2025.11
84.3
2025.11
84.3
2025.11
84.3
2025.11
83.9
2025.11
83.9
2025.11
83.1
2025.11
83.1
2025.11
83
2025.11
79.2
2025.11
77.6
2025.11
61
2025.11
61
2025.11
48.7
2025.11
44
2025.11
40.7
2025.11
40.3
2025.12
39.65
2025.12
39
2025.12
38.45
37.8
2025.11
37.7
2025.12
36.35
2025.12
36
2025.12
35.95
2025.12
35.63
2025.11
35.6
2025.12
22.22
2025.12
15.15
2026.02
12.5
2026.02
10.9
2026.02
10.9
2026.02
10.9
2026.02
10.9
2026.02
10.2
2026.02
10.2
2026.02
9.4
2026.02
8.6
2026.02
7
5.7
2025.12
5.49
2025.12
5.25
2025.12
5.1
2025.12
5
2025.12
4.5
2025.12
3.85
2025.12
3.8
2025.12
3.45
2025.12
3.3