Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Mathematical Reasoning on Minerva (Mean@32 accuracy)
Loading...
36.5
Mean@32 Accuracy
BAPO
27.244
29.647
32.05
34.453
May 8, 2026
Mean@32 Accuracy
Updated 2mo ago
Evaluation Results
Method
Method
Links
Mean@32 Accuracy
BAPO
Backbone=Qwen3-8B-Inst...
2026.05
36.5
HTPO
Backbone=Qwen3-8B-Inst...
2026.05
36.5
GSPO
Backbone=Qwen3-8B-Inst...
2026.05
35.9
GRPO†
Backbone=Qwen3-8B-Inst...
2026.05
34.8
80/20-Rule
Backbone=Qwen3-8B-Inst...
2026.05
34.3
DAPO
Backbone=Qwen3-8B-Inst...
2026.05
33.8
SAPO
Backbone=Qwen3-8B-Inst...
2026.05
33.1
GSPO
Backbone=Qwen3-8B-Base
2026.05
30.3
HTPO
Backbone=Qwen3-8B-Base
2026.05
30.3
SAPO
Backbone=Qwen3-8B-Base
2026.05
30.1
DAPO
Backbone=Qwen3-8B-Base
2026.05
29.4
GRPO†
Backbone=Qwen3-8B-Base
2026.05
28.5
BAPO
Backbone=Qwen3-8B-Base
2026.05
28.3
80/20-Rule
Backbone=Qwen3-8B-Base
2026.05
27.6
Feedback
Search any
task
Search any
task
