task,engine,model,scope,concurrency,planned_unique,completed_unique,correct_count,exact_packets,completion_rate,accuracy,macro_f1,packet_exact_match,page_accuracy,page_macro_f1,coverage_validity,coverage_valid_packets,same_category_boundary_recall,same_category_boundaries_found,same_category_boundaries_actual,boundary_precision,boundary_recall,boundary_f1,boundary_true_positive,boundary_predicted,boundary_actual,segment_precision,segment_recall,segment_f1,segment_true_positive,segment_predicted,segment_actual,planned_calls,dispatched_calls,successful_calls,failed_calls,skipped_calls,missing_terminal_calls,decision_p50_ms,decision_p95_ms,wall_p50_ms,wall_p95_ms,latency_samples,decision_cost_usd_known,unknown_cost_calls,request_count,request_count_complete classify,jev,jev-1.13.0,decision,1,40,40,40,,1.0,1.0,1.0,,,,,,,,,,,,,,,,,,,,,40,40,40,0,0,0,138.565,184.54974999999996,138.86691629886627,184.82249404769388,40,0.005028114,0,40,True classify,openai,gpt-5.6-luna,decision,1,40,40,40,,1.0,1.0,1.0,,,,,,,,,,,,,,,,,,,,,40,40,40,0,0,0,794.309,1208.7996999999998,794.5054790470749,1208.9875228237358,40,0.02440605,0,40,True split,jev,jev-1.13.0,decision,1,8,8,,7,1.0,,,0.875,1.0,1.0,1.0,8,1.0,4,4,0.9696969696969697,1.0,0.9846153846153847,32,33,32,0.9512195121951219,0.975,0.9629629629629629,39,41,40,8,8,8,0,0,0,209.578,289.98035,210.2170626167208,290.3458559885621,8,0.0066345300000000005,0,8,True split,openai,gpt-5.6-luna,decision,1,8,8,,8,1.0,,,1.0,1.0,1.0,1.0,8,1.0,4,4,1.0,1.0,1.0,32,32,32,1.0,1.0,1.0,40,40,40,8,8,8,0,0,0,1352.3175,1862.7393,1352.6357288938016,1863.4103753836825,8,0.022487649999999998,0,8,True