Paraphrase Identification on MCEval PAWS 8K (test)
88.9AccuracyAct
Evaluation Results
| Method | Links | |
|---|---|---|
| ActModel=Qwen2.5-7B2024.12 | 88.9 | |
| Magn-ProbeModel=Qwen2.5-7B2024.12 | 87.3 | |
| ActModel=Llama3.1-8B2024.12 | 86 | |
| Magn-ProbeModel=Llama3.1-8B2024.12 | 85.4 | |
| ActModel=Llama3.2-3B2024.12 | 84.3 | |
| Magn-ProbeModel=Llama3.2-3B2024.12 | 83.1 | |
| ActModel=Llama2-7B2024.12 | 82.5 | |
| Magn-ProbeModel=Llama2-70B2024.12 | 82.5 | |
| Magn-ProbeModel=Llama2-7B2024.12 | 81.5 | |
| LM-ProbModel=Llama2-70B2024.12 | 77.9 | |
| LM-ProbModel=Qwen2.5-7B2024.12 | 57.4 | |
| LM-ProbModel=Llama2-7B2024.12 | 52.4 | |
| LM-ProbModel=Llama3.2-3B2024.12 | 50 | |
| LM-ProbModel=Llama3.1-8B2024.12 | 50 |