ResearchBenchmarksMath Reasoning on GPQA (BA@20, BA@40, BA@60)Follow37.2BA @20RoRo34.635.27535.9536.625May 28, 2026Evaluation ResultsMethodMethodLinksBA @20BA @40BA @60RoRoSetting=Same-family, M...Setting=Same-family, Models=Qwen3-1.7B / Qwen3-14B2026.0537.244.648.6RoRoSetting=Cross-family,...Setting=Cross-family, Models=Qwen3-1.7B / DeepSeek-R1-Distill-Qwen-14B2026.0534.741.846.1