Paraphrase Identification on MRPC
-4.73Delta 1Qwen2.5-14B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen2.5-14Bscenario=updated dataset (ours)2025.11 | -4.73 | 0.37 | |
| Yi1.5-9Bscenario=updated dataset (ours)2025.11 | -3.5 | 0.27 | |
| Yi1.5-6Bscenario=updated dataset (ours)2025.11 | -1 | 0.21 | |
| Qwen2.5-7Bscenario=updated dataset (ours)2025.11 | 0.04 | -0.39 | |
| Mistral-12Bscenario=updated dataset (ours)2025.11 | 0.61 | 0.43 | |
| Llama3-8Bscenario=updated dataset (ours)2025.11 | 0.75 | 0.53 | |
| Qwen2.5-7Bscenario=semantic dataset2025.11 | 3.21 | 2.32 | |
| Ministral-8Bscenario=updated dataset (ours)2025.11 | 3.86 | 0.32 | |
| Qwen2.5-14Bscenario=semantic dataset2025.11 | 5.49 | 1.42 | |
| Llama3-8Bscenario=semantic dataset2025.11 | 6.37 | 2.78 | |
| Yi1.5-9Bscenario=semantic dataset2025.11 | 6.48 | 1.89 | |
| Yi1.5-6Bscenario=semantic dataset2025.11 | 6.92 | 1.39 | |
| Qwen2.5-7Bscenario=original dataset2025.11 | 8.06 | 4.16 | |
| Llama2-13Bscenario=updated dataset (ours)2025.11 | 10.09 | -0.31 | |
| Mistral-12Bscenario=semantic dataset2025.11 | 10.54 | 2.28 | |
| Llama3-8Bscenario=original dataset2025.11 | 10.98 | 7.05 | |
| Ministral-8Bscenario=semantic dataset2025.11 | 11.94 | 2.84 | |
| Yi1.5-6Bscenario=original dataset2025.11 | 14.96 | 7.71 | |
| Llama2-13Bscenario=semantic dataset2025.11 | 18.44 | 2.5 | |
| Qwen2.5-14Bscenario=original dataset2025.11 | 19.95 | 6.94 | |
| Yi1.5-9Bscenario=original dataset2025.11 | 23.87 | 10.41 | |
| Ministral-8Bscenario=original dataset2025.11 | 25.64 | 8.91 | |
| Mistral-12Bscenario=original dataset2025.11 | 26.51 | 9.3 | |
| Llama2-13Bscenario=original dataset2025.11 | 26.98 | 6.83 |