Response Generation with Provenance on ReFInE
57.25ROUGE-LGenProve
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| GenProve2026.01 | 57.25 | 42.22 | 59.39 | 51.04 | 54.96 | 51.26 | 51.21 | 99.85 | 3.14 | |
| Qwen3-14B2026.01 | 52.85 | 35.7 | 55.34 | 47.06 | 45.8 | 40.33 | 41.16 | 99.7 | 2.59 | |
| Qwen3-8B2026.01 | 51.8 | 35.56 | 55.38 | 45.9 | 37.78 | 30.56 | 32.53 | 100 | 2.25 | |
| GLM-4-9B2026.01 | 50.33 | 32.62 | 50 | 44.93 | 34.57 | 34.12 | 32.79 | 100 | 2.16 | |
| GLM-4.5-355B2026.01 | 49.81 | 35.05 | 57.69 | 44.92 | 48.84 | 44.03 | 44.55 | 98.63 | 2.63 | |
| Kimi2026.01 | 49.33 | 31.24 | 51.56 | 43.84 | 31.55 | 32.09 | 29.7 | 99.09 | 2.2 | |
| Gemma-3-12B2026.01 | 48.97 | 32.13 | 50.8 | 43.79 | 41.06 | 31.11 | 34.03 | 100 | 2.47 | |
| Gemini 2.5 Pro2026.01 | 48.75 | 31.77 | 53.09 | 44.79 | 46.68 | 42.86 | 42.92 | 100 | 2.57 | |
| Yi-1.5-9B2026.01 | 48.26 | 30.11 | 47.77 | 43.35 | 20.71 | 22.91 | 20.11 | 96.81 | 1.8 | |
| InternLM2.5-7B2026.01 | 47.79 | 30.6 | 53.47 | 43.6 | 12.64 | 13.35 | 11.94 | 80.24 | 1.67 | |
| Llama-3.1-8B2026.01 | 47.71 | 26.36 | 42.04 | 41.14 | 21.78 | 20.3 | 20.05 | 99.85 | 2 | |
| GPT-52026.01 | 41.79 | 20.01 | 38.83 | 36.38 | 21.37 | 16.73 | 17.88 | 99.85 | 2.23 | |
| Hunyuan-7B2026.01 | 39.91 | 24.7 | 43.74 | 32.99 | 22.78 | 22.86 | 21.43 | 90.88 | 1.72 | |
| Vicuna-7b-v1.52026.01 | 38.83 | 24.37 | 50.08 | 34.73 | 9.01 | 5.85 | 6.7 | 92.4 | 1.1 | |
| Baichuan2-7B2026.01 | 34.68 | 19.41 | 48.08 | 32.03 | 3.22 | 3.41 | 3.04 | 26.6 | 0.78 |