Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Mathematical Reasoning on MATH500 (mean@8)
Loading...
84.6
Mean@8 Score
ACPO
71.3608
74.7979
78.235
81.6721
Jul 3, 2026
Mean@8 Score
Updated 18d ago
Evaluation Results
Method
Method
Links
Mean@8 Score
ACPO
Model=DeepSeek-R1-Dist...
2026.07
84.6
ACPO
Model=Qwen3-8B-Base
2026.07
83.6
SAPO
Model=DeepSeek-R1-Dist...
2026.07
83.2
SAPO
Model=Qwen3-8B-Base
2026.07
82.4
GTPO
Model=Qwen3-8B-Base
2026.07
81.62
DAPO
Model=DeepSeek-R1-Dist...
2026.07
80.4
GTPO
Model=DeepSeek-R1-Dist...
2026.07
78.85
80/20
Model=DeepSeek-R1-Dist...
2026.07
78.72
80/20
Model=Qwen3-8B-Base
2026.07
78.22
GRPO
Model=DeepSeek-R1-Dist...
2026.07
77.95
ACPO
Model=Qwen2.5-Math-7B
2026.07
77.8
DAPO
Model=Qwen3-8B-Base
2026.07
76.8
SAPO
Model=Qwen2.5-Math-7B
2026.07
76.4
GTPO
Model=Qwen2.5-Math-7B
2026.07
75.5
DAPO
Model=Qwen2.5-Math-7B
2026.07
75
80/20
Model=Qwen2.5-Math-7B
2026.07
74.75
GRPO
Model=Qwen3-8B-Base
2026.07
72.8
GRPO
Model=Qwen2.5-Math-7B
2026.07
71.87
Feedback
Search any
task
Search any
task