List your company

Code Generation on MBPP+ (Pass@1, Pass@k)

69.44Pass@1ASFT
23.326435.298247.2759.2418Apr 1, 2026Apr 17, 2026May 3, 2026May 19, 2026Jun 4, 2026Jun 20, 2026Jul 6, 2026

Evaluation results

MethodLinks
ASFTModel=Qwen3-14B2026.0769.4477.81
EAFTModel=Qwen3-14B2026.0768.8881.03
CEModel=Qwen3-14B2026.0768.4379.92
LP-SFTModel=Qwen3-14B2026.0768.283.73
PriFT-probBackbone=Qwen3-4B-Instruct2026.0667.99—
EntropyModel=WeDLM-8B, Default decoding=true2026.0467.773.9
GEMModel=Qwen3-14B2026.0767.5982.46
DFTModel=Qwen3-14B2026.0766.4269.67
IMHModel=WeDLM-8B2026.0466.483.6
OriginalBackbone=Qwen3-4B-Instruct2026.0665.61—
TALRBackbone=Qwen3-4B-Instruct2026.0665.34—
PriFT-massBackbone=Qwen3-4B-Instruct2026.0664.81—
ASFTBackbone=Qwen3-4B-Instruct2026.0664.29—
EAFTBackbone=Qwen3-4B-Instruct2026.0663.76—
SFTBackbone=Qwen3-4B-Instruct2026.0663.23—
IDFTBackbone=Qwen3-4B-Instruct2026.0661.9—
ASFTModel=Qwen3-4B2026.0761.671.12
EAFTModel=Qwen3-4B2026.0761.273.55
DFTBackbone=Qwen3-4B-Instruct2026.0660.85—
CEModel=Qwen3-4B2026.0760.4272.4
GEMModel=Qwen3-4B2026.0760.1976.21
LP-SFTModel=Qwen3-4B2026.0760.1978.93
ConfidenceModel=WeDLM-8B2026.0459.871.7
DFTModel=Qwen3-4B2026.0758.9962.68
MarginModel=WeDLM-8B2026.044964.3
CEModel=Llama3.1-8B2026.0746.9262.31
LP-SFTModel=Llama3.1-8B2026.0746.5266.91
GEMModel=Llama3.1-8B2026.0746.4965.39
ASFTModel=Llama3.1-8B2026.0743.3657.21
EAFTModel=Llama3.1-8B2026.0742.0160.73
ConfidenceModel=LLaDA-8B, Default decoding=true2026.0440.452.1
IMHModel=LLaDA-8B2026.0439.967.2
MarginModel=LLaDA-8B2026.0438.858.9
EntropyModel=LLaDA-8B2026.0438.765.6
DFTModel=Llama3.1-8B2026.0730.735.98
RandomModel=LLaDA-8B2026.0425.160.5