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

Mathematical Reasoning on MATH 2021 (test)

78.3pass@1

FedAvg-GRPO

54.27660.51366.7572.987Apr 14, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.04
78.3
2026.04
77.5
2026.04
77
2026.04
76.9
2026.04
76.9
2026.04
76.7
2026.04
76.6
2026.04
76.5
2026.04
76.3
2026.04
76.1
2026.04
75.9
2026.04
75.6
2026.04
73.7
2026.04
73.5
2026.04
73.2
2026.04
73.1
2026.04
73.1
2026.04
72.8
2026.04
71.9
2026.04
71.9
2026.04
71.5
2026.04
71.4
2026.04
71.1
2026.04
70.6
2026.04
58.4
2026.04
58.4
2026.04
58.4
2026.04
58.4
2026.04
55.2
2026.04
55.2
2026.04
55.2
2026.04
55.2