Commonsense Multi-hop QA on StrategyQA
8.1Full Length CoverageMinimal-core extraction
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Minimal-core extractionModel=GPT-52026.05 | 8.1 | 3.2 | 40 | 60 | 80 | 95 | |
| Minimal-core extractionModel=DeepSeek-R1-Distill-Qwen-32B2026.05 | 7.8 | 3.4 | 44 | 56 | 76 | 93 | |
| Minimal-core extractionModel=Qwen3-32B2026.05 | 7.5 | 3.5 | 47 | 53 | 73 | 91 | |
| Minimal-core extractionModel=DeepSeek-R1-Distill-Qwen-7B2026.05 | 7.4 | 3.7 | 50 | 50 | 69 | 88 |