Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Mathematical Reasoning on AIME 25 (A25 Score)
Loading...
62.3
A25 Score
GEAR
33.596
41.048
48.5
55.952
May 12, 2026
A25 Score
Updated 21d ago
Evaluation Results
Method
Method
Links
A25 Score
GEAR
Backbone=Qwen3-4B, Eva...
2026.05
62.3
GEAR
Backbone=Qwen3-8B, Eva...
2026.05
54.2
GEAR
Backbone=Qwen3-4B, Eva...
2026.05
53.4
GRPO
Backbone=Qwen3-4B, Eva...
2026.05
52.5
Base
Backbone=Qwen3-8B, Eva...
2026.05
52.3
Base
Backbone=Qwen3-4B, Eva...
2026.05
51.2
ARPO
Backbone=Qwen3-4B, Eva...
2026.05
48.9
ARPO
Backbone=Qwen3-8B, Eva...
2026.05
47.8
GRPO
Backbone=Qwen3-4B, Eva...
2026.05
47.4
GRPO
Backbone=Qwen3-8B, Eva...
2026.05
46.3
OPSD+RL
Backbone=Qwen3-4B, Eva...
2026.05
45.8
MT-GRPO
Backbone=Qwen3-4B, Eva...
2026.05
45.6
MT-GRPO
Backbone=Qwen3-8B, Eva...
2026.05
44.6
OPSD+RL
Backbone=Qwen3-8B, Eva...
2026.05
44.5
OPSD
Backbone=Qwen3-4B, Eva...
2026.05
42.1
OPSD
Backbone=Qwen3-8B, Eva...
2026.05
34.7
Feedback
Search any
task
Search any
task