Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
General Reasoning on GSM8k, MATH 500, GPQA, SuperGPQA
Loading...
63.8
Average Accuracy
BF16
6.288
21.219
36.15
51.081
Jan 20, 2026
Average Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Average Accuracy
BF16
Model=Qwen3-8B-Base, R...
2026.01
63.8
Jet-RL
Model=Qwen3-8B-Base, R...
2026.01
62.7
BF16-Train-FP8-Rollout
Model=Qwen3-8B-Base, R...
2026.01
60.9
BF16
Model=Qwen2.5-7B, Roll...
2026.01
56.9
Jet-RL
Model=Qwen2.5-7B, Roll...
2026.01
55.9
Initial
Model=Qwen3-8B-Base, R...
2026.01
52.6
Initial
Model=Qwen2.5-7B, Roll...
2026.01
29.6
Jet-RL
Model=Llama3.1-8B, Rol...
2026.01
25.2
BF16
Model=Llama3.1-8B, Rol...
2026.01
23.2
BF16-Train-FP8-Rollout
Model=Llama3.1-8B, Rol...
2026.01
13
Initial
Model=Llama3.1-8B, Rol...
2026.01
8.5
Feedback
Search any
task
Search any
task