List your company

Code Generation on HumanEval 29 tasks

96.72Mean F1 Scorezebra-mult_offset
94.088894.771995.45596.1381May 19, 2026

Evaluation results

MethodLinks
zebra-mult_offsetalpha=0.5, runs_per_cell=152026.0596.7299.1110.0253
zebra-additivealpha=0.5, runs_per_cell=152026.0596.2698.690.0754
llm_cotalpha=0.5, runs_per_cell=152026.0594.639780.8752
llmalpha=0.5, runs_per_cell=152026.0594.1996.5——