ResearchBenchmarksScientific/Knowledge Reasoning on Scientific GPQA and MMLUFollow65.2ScoreDASD62.39263.12163.8564.579May 21, 2026Evaluation ResultsMethodMethodLinksScoreDASDBackbone=Qwen3-8BBackbone=Qwen3-8B2026.0565.2OPSDBackbone=Qwen3-8BBackbone=Qwen3-8B2026.0564.3GRPOBackbone=Qwen3-8BBackbone=Qwen3-8B2026.0563Qwen3-8BPost-training=BasePost-training=Base2026.0562.5