Hallucination Evaluation on MCEval HaluEval 8K (test)
82.2AccuracyAct
Evaluation Results
| Method | Links | |
|---|---|---|
| ActModel=Llama3.2-3B2024.12 | 82.2 | |
| ActModel=Qwen2.5-7B2024.12 | 82.1 | |
| ActModel=Llama3.1-8B2024.12 | 80.7 | |
| Magn-ProbeModel=Llama3.1-8B2024.12 | 79.8 | |
| ActModel=Llama2-7B2024.12 | 78.8 | |
| Magn-ProbeModel=Llama2-7B2024.12 | 78.3 | |
| Magn-ProbeModel=Llama3.2-3B2024.12 | 78 | |
| Magn-ProbeModel=Qwen2.5-7B2024.12 | 77.3 | |
| Magn-ProbeModel=Llama2-70B2024.12 | 76.3 | |
| LM-ProbModel=Llama2-70B2024.12 | 75.4 | |
| LM-ProbModel=Llama3.1-8B2024.12 | 67.4 | |
| LM-ProbModel=Llama3.2-3B2024.12 | 65 | |
| LM-ProbModel=Qwen2.5-7B2024.12 | 63.7 | |
| LM-ProbModel=Llama2-7B2024.12 | 52 |