List your company

Code Reasoning on LiveCodeBench 1.0 (test)

47.2AccuracyA3PO
10.17619.78829.439.012Dec 25, 2025

Evaluation results

MethodLinks
A3POModel=Deepseek-R1-Distill-Qwen-7B2025.1247.2
Lp-RegModel=Deepseek-R1-Distill-Qwen-7B2025.1244.7
W-REINFORCEModel=Deepseek-R1-Distill-Qwen-7B2025.1244.6
DAPO w/ Fork TokensModel=Deepseek-R1-Distill-Qwen-7B2025.1244.1
DAPOModel=Deepseek-R1-Distill-Qwen-7B2025.1243.2
GRPOModel=Deepseek-R1-Distill-Qwen-7B2025.1242.5
A3POModel=Qwen3-8B-Base2025.1233.8
DAPO w/ Fork TokensModel=Qwen3-8B-Base2025.1231.2
Lp-RegModel=Qwen3-8B-Base2025.1230.9
W-REINFORCEModel=Qwen3-8B-Base2025.1230.4
DAPOModel=Qwen3-8B-Base2025.1229.7
GRPOModel=Qwen3-8B-Base2025.1229.4
A3POModel=Qwen2.5-7B-Math2025.1216.4
DAPO w/ Fork TokensModel=Qwen2.5-7B-Math2025.1214.3
W-REINFORCEModel=Qwen2.5-7B-Math2025.1213.8
Lp-RegModel=Qwen2.5-7B-Math2025.1213.8
DAPOModel=Qwen2.5-7B-Math2025.1212.4
GRPOModel=Qwen2.5-7B-Math2025.1211.6