ResearchBenchmarksMathematical reasoning on AMC 23 (pass@1, pass@64)Follow20.5Pass@1DSTaR13.989615.679817.3719.0602May 8, 2026Evaluation ResultsMethodMethodLinksPass@1Pass@64DSTaRn=N/An=N/A2026.0520.586.75DPólyan=32n=322026.0517.8485.18DPólyan=64n=642026.0517.6485.01DPólyan=16n=162026.0517.5584.03Zero-Shotn=N/An=N/A2026.0516.9583.49DPólyan=8n=82026.0516.7983.19DPólyan=4n=42026.0516.3683.51DPólyan=2n=22026.0516.0884.45DPólyan=1n=12026.0514.2482.79