2026-10-22-o1-sampling-chart
claim_id,model,evaluation,candidates_per_problem,selection_method,accuracy_percent,precision,reported_average_correct_per_15,usd_cost_per_problem,source_locator,source_id,retrieved_date
UP03,o1 research model,AIME 2024,1,single_sample,74.4,appendix_one_decimal,,,Appendix A pass@1,UP-S04,2026-10-22
UP03,o1 research model,AIME 2024,64,consensus,83.3,appendix_one_decimal,,,Appendix A cons@64,UP-S04,2026-10-22
UP03-BACKUP,o1 research model,AIME 2024,1000,learned_reranking,93,main_text_rounded,13.9,,Evals main text,UP-S04,2026-10-22