Multi-hop Question Answering on 2WikiMultiHopQA (LLM Judge)
81.53LLM Judge ScoreQwen3-1.7B-SFT
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-1.7B-SFTBackbone=GLM-5.1, Evaluation Protocol=mean@42026.07 | 81.53 | |
| DeepSeek-V4-FlashBackbone=GLM-5.1, Evaluation Protocol=mean@42026.07 | 81.33 | |
| Qwen3-1.7B-SFTBackbone=DeepSeek-V4-Flash, Evaluation Protocol=mean@42026.07 | 81.13 | |
| DeepSeek-V4-FlashBackbone=DeepSeek-V4-Flash, Evaluation Protocol=mean@42026.07 | 80.73 | |
| Qwen3-1.7BBackbone=GLM-5.1, Evaluation Protocol=mean@42026.07 | 75.96 | |
| Qwen3-1.7BBackbone=DeepSeek-V4-Flash, Evaluation Protocol=mean@42026.07 | 75.38 |