Multi-hop Question Quality Evaluation on Comparison Questions
98.6Multi-Hop AccuracyHopWeaver
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| HopWeaverLLM Engine=Gemini-2.5-flash2025.05 | 98.6 | 4.45 | |
| 2WikiMultiHopQAGeneration Source=Human Datasets (Baselines)2025.05 | 97.6 | 4.42 | |
| HopWeaverLLM Engine=QwQ-32B2025.05 | 97.4 | 4.4 | |
| HopWeaverLLM Engine=Qwen3-14B2025.05 | 95.9 | 4.36 | |
| HotpotQAGeneration Source=Human Datasets (Baselines)2025.05 | 95.6 | 4.2 | |
| HopWeaverLLM Engine=GLM-4-9B-04142025.05 | 93.9 | 4.26 |