ResearchBenchmarksMedical Question Answering on LancetFollow70.15Accuracym1-32B-1K62.318864.351966.38568.4181Apr 1, 2025Evaluation ResultsMethodMethodLinksAccuracym1-32B-1KParameter Scale=32B, T...Parameter Scale=32B, Thinking Budget=1K2025.0470.15m1-7B-23KParameter Scale=7B, Th...Parameter Scale=7B, Thinking Budget=23K2025.0462.62