ResearchBenchmarksMathematical Reasoning on AIME24 (pass@8)Follow83.33Pass@8GRPO55.603662.80187077.1982May 9, 2026Evaluation ResultsMethodMethodLinksPass@8GRPO2026.0583.33OPSD2026.0580OPHSDharness=plan-solveharness=plan-solve2026.0580OPHSDharness=noneharness=none2026.0579.17CRISP2026.0556.67