Multi-hop Question Answering on WebQSP
68Pass@1o3-mini
Evaluation Results
| Method | Links | |
|---|---|---|
| o3-mini2025.05 | 68 | |
| Qwen2.5-72B2025.05 | 65.3 | |
| Qwen2.5-32B2025.05 | 62.1 | |
| Qwen2.5-32B + cotProtocol=CoT2025.05 | 61.4 | |
| QwQ-32B2025.05 | 61.2 | |
| Qwen2.5-14B2025.05 | 59.9 | |
| Qwen2.5-14B + cotProtocol=CoT2025.05 | 59.2 | |
| Qwen2.5-7B2025.05 | 58 | |
| Qwen2.5-32B + fs1Protocol=fs12025.05 | 57.6 | |
| Qwen2.5-7B + cotProtocol=CoT2025.05 | 56.5 | |
| Qwen2.5-32B + rtProtocol=rt2025.05 | 56.3 | |
| Qwen2.5-14B + rtProtocol=rt2025.05 | 56.2 | |
| Qwen2.5-14B + fs1Protocol=fs12025.05 | 55.8 | |
| R1-70B2025.05 | 54.9 | |
| Qwen2.5-7B + rtProtocol=rt2025.05 | 51.7 | |
| Qwen2.5-7B + fs1Protocol=fs12025.05 | 49.2 | |
| Qwen2.5-3B + rtProtocol=rt2025.05 | 48.3 | |
| Qwen2.5-3B + cotProtocol=CoT2025.05 | 47.7 | |
| Qwen2.5-3B2025.05 | 46.6 | |
| SmolLM2-1.7B + rtProtocol=rt2025.05 | 45.4 | |
| SmolLM2-1.7B + fs1Protocol=fs12025.05 | 42.6 | |
| SmolLM2-1.7B + cotProtocol=CoT2025.05 | 40.9 | |
| SmolLM2-1.7B2025.05 | 40.8 | |
| Qwen2.5-3B + fs1Protocol=fs12025.05 | 40.6 | |
| Qwen2.5-1.5B + cotProtocol=CoT2025.05 | 39.1 | |
| Qwen2.5-1.5B2025.05 | 36 | |
| Qwen2.5-1.5B + rtProtocol=rt2025.05 | 36 | |
| Qwen2.5-1.5B + fs1Protocol=fs12025.05 | 35.6 | |
| Qwen2.5-0.5B + fs1Protocol=fs12025.05 | 30.2 | |
| SmolLM2-360M + rtProtocol=rt2025.05 | 29.3 | |
| Qwen2.5-0.5B + rtProtocol=rt2025.05 | 28.6 | |
| SmolLM2-360M + fs1Protocol=fs12025.05 | 26.4 | |
| SmolLM2-360M2025.05 | 23.5 | |
| Qwen2.5-0.5B + cotProtocol=CoT2025.05 | 23.4 | |
| SmolLM2-360M + cotProtocol=CoT2025.05 | 23 | |
| Qwen2.5-0.5B2025.05 | 17.3 |