List your company

Code Generation on MBPP (Pass%, CodeBLEU)

99.78Pass%Iterative
41.59256.698571.80586.9115Jun 18, 2026

Evaluation results

MethodLinks
IterativeModels=Gemini2026.0699.781.20.97470.789.10.46
Refactoring-BasedModels=Claude2026.0699.67—0.723880.65
Refactoring-BasedModels=Gemini2026.0699.45—0.87911.142.20.91
StaticModels=Gemini2026.0698.25—0.92239.859.70.44
IterativeModels=Claude2026.0697.271.330.98990960.46
Iterative-AGLModels=Gemini2026.0696.791.90.99496.797.90.46
Comment-basedModels=Gemini2026.0695.01—0.92541.868.60.44
ICWModels=Gemini2026.0694.44—0.5124.57.20.44
Iterative-AGLModels=Claude2026.0693.62.50.98787940.47
ICWModels=Claude2026.0692.4—0.772140.5
Comment-basedModels=Claude2026.0690—0.89239600.47
StaticModels=Claude2026.0686.73—0.89926560.46
No WatermarkModels=Gemini2026.0664.77—0.5773.510.60.42
No WatermarkModels=Claude2026.0663.96—0.416460.5
SWEETModels=Qwen2026.0643.83—0.9244720.42