Open-ended generation on MATH L5
0.35FDRSGR
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| SGRModel=Qwen3-4B-Thinking-2507, Uncertainty Score=logits-based2025.10 | 0.35 | 1.96 | 1.95 | — | |
| SGRModel=Qwen3-4B-Instruct-2507, Uncertainty Score=logits-based2025.10 | 0.55 | 2.9 | 2.85 | — | |
| SGRModel=DeepSeek-R1-Distill-Qwen-32B, Uncertainty Score=logits-based2025.10 | 1.42 | 13.86 | 13.71 | — | |
| Conformal LabelingModel=Qwen3-4B-Thinking-2507, Uncertainty Score=logits-based2025.10 | 4.64 | 48.33 | 48.45 | — | |
| Conformal LabelingModel=DeepSeek-R1-Distill-Qwen-32B, Uncertainty Score=logits-based2025.10 | 4.69 | 65.31 | 64.69 | — | |
| FDR searchModel=Qwen3-4B-Thinking-2507, Uncertainty Score=logits-based2025.10 | 4.72 | 57.85 | 57.94 | — | |
| Conformal LabelingModel=Qwen3-4B-Instruct-2507, Uncertainty Score=logits-based2025.10 | 4.84 | 69.54 | 66.9 | — | |
| FDR searchModel=Qwen3-4B-Instruct-2507, Uncertainty Score=logits-based2025.10 | 5.14 | 68.89 | 66.25 | — | |
| FDR searchModel=DeepSeek-R1-Distill-Qwen-32B, Uncertainty Score=logits-based2025.10 | 6.1 | 68.66 | 68.04 | — | |
| AI onlyModel=Qwen3-4B-Instruct-25072025.10 | — | — | — | 8.88 | |
| AI onlyModel=Qwen3-4B-Thinking-25072025.10 | — | — | — | 5.27 | |
| AI onlyModel=DeepSeek-R1-Distill-Qwen-32B2025.10 | — | — | — | 6.38 |