ResearchBenchmarksMathematical Reasoning on HMMT25 (pass@8)Follow53.33Pass@8OPHSD42.066844.990947.91550.8391May 9, 2026Evaluation ResultsMethodMethodLinksPass@8OPHSDharness=plan-solveharness=plan-solve2026.0553.33OPHSDharness=noneharness=none2026.0553.33GRPO2026.0545CRISP2026.0543.33OPSD2026.0542.5