Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Mathematical Reasoning on BeyondAIME (mean@8)
Loading...
21
Mean@8 Score
ACPO
3.32
7.91
12.5
17.09
Jul 3, 2026
Mean@8 Score
Updated 18d ago
Evaluation Results
Method
Method
Links
Mean@8 Score
ACPO
Model=DeepSeek-R1-Dist...
2026.07
21
GTPO
Model=DeepSeek-R1-Dist...
2026.07
19.12
SAPO
Model=DeepSeek-R1-Dist...
2026.07
18.62
80/20
Model=DeepSeek-R1-Dist...
2026.07
15.75
ACPO
Model=Qwen3-8B-Base
2026.07
15.5
DAPO
Model=DeepSeek-R1-Dist...
2026.07
15.12
GRPO
Model=DeepSeek-R1-Dist...
2026.07
13.62
GTPO
Model=Qwen3-8B-Base
2026.07
12.62
SAPO
Model=Qwen3-8B-Base
2026.07
12.25
80/20
Model=Qwen3-8B-Base
2026.07
12
DAPO
Model=Qwen3-8B-Base
2026.07
11.75
GRPO
Model=Qwen3-8B-Base
2026.07
9.5
ACPO
Model=Qwen2.5-Math-7B
2026.07
8
GTPO
Model=Qwen2.5-Math-7B
2026.07
7.12
SAPO
Model=Qwen2.5-Math-7B
2026.07
6.75
80/20
Model=Qwen2.5-Math-7B
2026.07
5.25
DAPO
Model=Qwen2.5-Math-7B
2026.07
4.5
GRPO
Model=Qwen2.5-Math-7B
2026.07
4
Feedback
Search any
task
Search any
task
