Hallucination Detection on RAGTruth Llama-7B
92.54RecallSelfCheckGPT
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SelfCheckGPTEvaluator Model=GPT-3.5-Turbo2026.01 | 92.54 | 53.27 | 67.62 | |
| SelfCheckGPTEvaluator Model=Llama-7B2026.01 | 83.53 | 53.32 | 65.09 | |
| SIRG2026.01 | 79.83 | 73.64 | 76.61 | |
| PromptEvaluator Model=Llama-7B2026.01 | 76.08 | 52.64 | 62.23 | |
| LRP4RAG2026.01 | 75.78 | 71.18 | 73.54 | |
| SEP2026.01 | 74.77 | — | 66.27 | |
| EigenScore2026.01 | 74.69 | — | 66.82 | |
| Fine-tuningEvaluator Model=Llama-7B2026.01 | 65.75 | 62.5 | 63.58 | |
| Fine-tuningEvaluator Model=Qwen2-7B2026.01 | 64.34 | 61.76 | 61.9 | |
| PromptEvaluator Model=GPT-3.5-Turbo2026.01 | 47.65 | 56.91 | 51.87 |