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

Mathematical Reasoning on AIME 26

86.7Score

Gpt-oss-20b-high

-3.46819.94143.3566.759Mar 17, 2026Mar 20, 2026Mar 24, 2026Mar 28, 2026Apr 1, 2026Apr 5, 2026Apr 9, 2026
Updated 5d ago

Evaluation Results

MethodLinks
2026.03
86.7
2026.03
86.7
2026.03
86.7
2026.03
70
2026.04
62.24
2026.04
57.24
2026.04
55.68
2026.04
55.21
2026.04
54.58
2026.04
54.38
2026.04
54.22
2026.04
53.91
2026.04
53.49
2026.04
51.25
2026.04
37.6
2026.03
36.7
2026.04
36.15
2026.04
35.99
2026.04
35.89
2026.04
35.63
2026.04
35.57
2026.04
35.52
2026.04
35.26
2026.04
35.1
2026.04
33.23
2026.03
10
2026.03
0.1
2026.03
0