List your company

Reasoning on PUPA (test)

91.85ScoreGEPA
80.378883.356986.33589.3131Jul 25, 2025

Evaluation results

MethodLinks
GEPABackbone=Qwen3 8B, Optimization Budget (# rollouts)=24262025.0791.85
GRPOBackbone=Qwen3 8B, Optimization Budget (# rollouts)=240002025.0786.66
GEPA+MergeBackbone=Qwen3 8B, Optimization Budget (# rollouts)=24262025.0786.26
MIPROv2Backbone=Qwen3 8B2025.0781.55
BaselineBackbone=Qwen3 8B2025.0780.82