ResearchBenchmarksMathematical Reasoning on AIME 24 (pass@1, pass@64)Follow4.39Pass@1DSTaR1.49882.249433.7506May 8, 2026Evaluation ResultsMethodMethodLinksPass@1Pass@64DSTaRn=N/An=N/A2026.054.3930.44Zero-Shotn=N/An=N/A2026.052.8132.08DPólyan=32n=322026.052.6433.27DPólyan=64n=642026.052.4832.32DPólyan=16n=162026.052.3532.06DPólyan=8n=82026.051.9931.2DPólyan=4n=42026.051.8331.04DPólyan=2n=22026.051.8231.42DPólyan=1n=12026.051.6130.72