Science Question Answering on GPQA Diamond (avg@1)
61.62Avg@1 ScoreSTAR-1-mix
Evaluation Results
| Method | Links | |
|---|---|---|
| STAR-1-mixModel=Qwen3-14B, Setting=STAR-1-mix2026.02 | 61.62 | |
| AgentModel=Qwen3-14B, Setting=Agent2026.02 | 60.61 | |
| AgentModel=Qwen3-8B, Setting=Agent2026.02 | 59.6 | |
| IFPOModel=Qwen3-14B, Setting=IFPO2026.02 | 58.59 | |
| STAR-1Model=Qwen3-14B, Setting=STAR-12026.02 | 58.08 | |
| BCModel=Qwen3-14B, Setting=BC2026.02 | 56.57 | |
| BCModel=Qwen3-8B, Setting=BC2026.02 | 54.55 | |
| IFPOModel=Qwen3-8B, Setting=IFPO2026.02 | 53.03 | |
| IFPOModel=Qwen3-4B, Setting=IFPO2026.02 | 52.53 | |
| STAR-1-mixModel=Qwen3-4B, Setting=STAR-1-mix2026.02 | 52.02 | |
| BCModel=Qwen3-4B, Setting=BC2026.02 | 51.52 | |
| STAR-1-mixModel=Qwen3-8B, Setting=STAR-1-mix2026.02 | 51.52 | |
| AgentModel=Qwen3-4B, Setting=Agent2026.02 | 50.51 | |
| STAR-1Model=Qwen3-4B, Setting=STAR-12026.02 | 49.49 | |
| SafeChainModel=Qwen3-14B, Setting=SafeChain2026.02 | 49.49 | |
| SafeChainModel=Qwen3-8B, Setting=SafeChain2026.02 | 47.98 | |
| STAR-1Model=Qwen3-8B, Setting=STAR-12026.02 | 47.47 | |
| SafeChainModel=Qwen3-4B, Setting=SafeChain2026.02 | 44.44 | |
| Trinity Large Baseshot=5-shot2026.02 | 43.94 |