ResearchBenchmarksMathematical Reasoning on AIME 25 (pass@1, pass@64)Follow44Pass@1DPólya25.2830.143539.86May 8, 2026Evaluation ResultsMethodMethodLinksPass@1Pass@64DPólyan=64n=642026.054418.66DPólyan=2n=22026.054017.44DPólyan=16n=162026.054017.52DPólyan=32n=322026.054016.97DPólyan=1n=12026.053816.71DPólyan=4n=42026.053716.7DPólyan=8n=82026.053415.26DSTaRn=N/An=N/A2026.053313.91Zero-Shotn=N/An=N/A2026.052612.84