List your company

Code Suggestion on DS-1000

51Binary RewardShapE-GRPO
39.35242.37645.448.424Mar 31, 2026

Evaluation results

MethodLinks
ShapE-GRPObackbone=QWEN3-8B2026.0351
GRPObackbone=QWEN3-8B2026.0346.5
WTAbackbone=QWEN3-8B2026.0341.8
QWEN3-8Bstatus=pretrained2026.0339.8