List your company

Code Optimization on HumanEval-Hard Cross-Domain

13.6Calls per TaskTextBFGS
12.71218.70624.730.694Jan 20, 2026

Evaluation results

MethodLinks
TextBFGSKnowledge Base=MBPP KB, Architecture=One-Pass2026.0113.61,581.721.6
TextBFGS-REMOKnowledge Base=MBPP KB, Architecture=One-Pass2026.0113.91,594.422.2
TextBFGSKnowledge Base=None2026.01171,481.325.2
TextGrad-Momentum2026.0129.81,464.243.7
TextGrad2026.0135.8863.930.9