List your company

Multi-agent Selection (Pairwise Resolution) on HumanEval 19 models (test)

80.9Pairwise ResolutionMARGIN
58.22864.1147075.886May 21, 2026

Evaluation results

MethodLinks
MARGINSelection Strategy=Verbalized2026.0580.9
Raw verbalized confidenceSelection Strategy=Raw verbalized2026.0559.1