ResearchBenchmarksGeneral Language Understanding on English Grouped General-Purpose MetricsFollow61.96Accuracy (English)Qwen3-8B57.820858.895459.9761.0446Sep 29, 2025Evaluation ResultsMethodMethodLinksAccuracy (English)Qwen3-8BScale=8BScale=8B2025.0961.968B-A-GRPOScale=8BScale=8B2025.0959.568B-GRPOScale=8BScale=8B2025.0957.98