Scientific Question Answering on GPQA Diamond (OOD) (test)
89.84AccuracySkillFlow
Evaluation Results
| Method | Links | |
|---|---|---|
| SkillFlowMethod Type=Ours, Backbone Model=Qwen3.5-9B2026.05 | 89.84 | |
| FlowSteerMethod Type=Agent+RL, Backbone Model=Qwen3.5-9B2026.05 | 84.38 | |
| AgentFlowMethod Type=Agent+RL, Backbone Model=Qwen3.5-9B2026.05 | 81.25 | |
| SkillRLMethod Type=Agent+RL, Backbone Model=Qwen3.5-9B2026.05 | 78.13 | |
| AFlowMethod Type=Search-based workflow, Backbone Model=Qwen3.5-9B2026.05 | 75.78 | |
| Qwen3.5Method Type=GRPO, Backbone Model=Qwen3.5-9B2026.05 | 75 | |
| v4-flashMethod Type=Baseline, Backbone Model=v4-flash2026.05 | 73.44 | |
| Qwen3.5Method Type=SFT, Backbone Model=Qwen3.5-9B2026.05 | 71.88 | |
| Qwen3.5Method Type=Baseline, Backbone Model=Qwen3.5-9B2026.05 | 61.72 |