List your company

Code Generation on BigCodeBench-I Hard

28.4ScoreGPT-o1
17.79220.54623.326.054Sep 26, 2025

Evaluation results

MethodLinks
GPT-o1Backbone=GPT-o12025.0928.4
DeepSeek-V2.5-238BBackbone=DeepSeek-V2.5-238B2025.0927
CRITIQUE-CODERBackbone=Qwen3-8B, Thinking Mode=true, Training Strategy=CRL2025.0927
Qwen3-8B-RLBackbone=Qwen3-8B, Thinking Mode=true, Training Strategy=Standard RL2025.0924.3
Baseline (Qwen3-8B)Backbone=Qwen3-8B, Thinking Mode=true, Training Strategy=Baseline2025.0923.6
Qwen3-4B-RLBackbone=Qwen3-4B, Thinking Mode=true, Training Strategy=Standard RL2025.0923
CRITIQUE-CODERBackbone=Qwen3-4B, Thinking Mode=true, Training Strategy=CRL2025.0923
DeepSeek-R1-Distill-14BBackbone=DeepSeek-R1-Distill-14B2025.0920.9
Baseline (Qwen3-4B)Backbone=Qwen3-4B, Thinking Mode=true, Training Strategy=Baseline2025.0920.9
AceCoder-7BBackbone=AceCoder-7B2025.0919.6
DeepCoder-14BBackbone=DeepCoder-14B2025.0918.2