Hallucination Detection on RAGTruth Llama2-13B (test)
83.33AccRAGLens
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| RAGLens2025.12 | 83.33 | 0.8964 | 81.48 | |
| ReDeEP2025.12 | 78.89 | 0.8244 | 75.87 | |
| ChainPoll2025.12 | 73.78 | 0.7414 | 73.7 | |
| Prompt2025.12 | 73 | — | 68.99 | |
| SEP2025.12 | 72.88 | 0.8098 | 77.99 | |
| RefCheck2025.12 | 72 | 0.7897 | 78.23 | |
| RAGAS2025.12 | 70.8 | 0.7541 | 69.87 | |
| Llama2-13B(LR)classifier=Logistic Regression, features=Llama2-13B2025.12 | 70.44 | — | 67.25 | |
| TurLens2025.12 | 67.56 | 0.7073 | 70.63 | |
| P(True)2025.12 | 62.66 | 0.8496 | 70.38 | |
| ITI2025.12 | 61.77 | 0.8501 | 68.5 | |
| Focus2025.12 | 60 | 0.7888 | 67.58 | |
| LwMLM2025.12 | 59.56 | — | 76.84 | |
| SelfCheckGPT2025.12 | 58.44 | — | 46.42 | |
| LN-Entropy2025.12 | 56.2 | 0.5912 | 68.5 | |
| SAPLMA2025.12 | 54.88 | 0.8029 | 69.23 | |
| Perplexity2025.12 | 53.33 | 0.5091 | 67.49 | |
| EigenScore2025.12 | 52.67 | 0.664 | 66.37 | |
| Energy2025.12 | 50.88 | 0.5619 | 66.57 | |
| FactCC2025.12 | 48 | 0.4753 | 41.21 | |
| FAcTScore2025.12 | 45.33 | 0.5294 | 62.39 |