Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Code Generation on LiveCodeBench (LCB) (% Avg@4)
Loading...
32.4
% Avg@4
RF++ B + RePro
17.424
21.312
25.2
29.088
Dec 1, 2025
% Avg@4
Updated 4d ago
Evaluation Results
Method
Method
Links
% Avg@4
RF++ B + RePro
Backbone=Qwen3-1.7B
2025.12
32.4
GRPO + RePro
Backbone=Qwen3-1.7B
2025.12
32
PPO + RePro
Backbone=Qwen3-1.7B
2025.12
31.5
PPO
Backbone=Qwen3-1.7B
2025.12
31.3
RF++ B
Backbone=Qwen3-1.7B
2025.12
31.1
Original
Backbone=Qwen3-1.7B
2025.12
30.6
GRPO
Backbone=Qwen3-1.7B
2025.12
30.2
GRPO + RePro
Backbone=Hunyuan-1.8B-...
2025.12
27.7
GRPO
Backbone=Hunyuan-1.8B-...
2025.12
27.1
RF++ B
Backbone=Hunyuan-1.8B-...
2025.12
26.5
PPO + RePro
Backbone=Hunyuan-1.8B-...
2025.12
26.3
RF++ B + RePro
Backbone=Hunyuan-1.8B-...
2025.12
26.2
PPO
Backbone=Hunyuan-1.8B-...
2025.12
25.9
Original
Backbone=Hunyuan-1.8B-...
2025.12
24.5
PPO + RePro
Backbone=MobileLLM-R1-...
2025.12
21.4
PPO
Backbone=MobileLLM-R1-...
2025.12
19.8
Original
Backbone=MobileLLM-R1-...
2025.12
18
Feedback
Search any
task
Search any
task