Hallucination Detection on SC-ChatGPT (AUCPR)
94.1AUCPRSELFCHECKGPT
Evaluation Results
| Method | Links | |
|---|---|---|
| SELFCHECKGPTDetector=LLaMA 3.1, Future context=true2025.07 | 94.1 | |
| SELFCHECKGPTDetector=Gemma 3, Future context=true2025.07 | 92.5 | |
| SELFCHECKGPTDetector=LLaMA 3.1, Future context=false2025.07 | 92 | |
| SELFCHECKGPTDetector=Gemma 3, Future context=false2025.07 | 91.7 | |
| DIRECTDetector=LLaMA 3.1, Future context=true2025.07 | 91.5 | |
| SCDetector=LLaMA 3.1, Future context=true2025.07 | 91.4 | |
| DIRECTDetector=Qwen 2.5, Future context=true2025.07 | 90.8 | |
| SELFCHECKGPTDetector=Qwen 2.5, Future context=true2025.07 | 90.6 | |
| SCDetector=Gemma 3, Future context=true2025.07 | 90.1 | |
| SELFCHECKGPTDetector=Qwen 2.5, Future context=false2025.07 | 90.1 | |
| DIRECTDetector=LLaMA 3.1, Future context=false2025.07 | 90 | |
| DIRECTDetector=Qwen 2.5, Future context=false2025.07 | 89.6 | |
| SCDetector=LLaMA 3.1, Future context=false2025.07 | 89.1 | |
| SCDetector=Gemma 3, Future context=false2025.07 | 88.6 | |
| SCDetector=Qwen 2.5, Future context=true2025.07 | 88.4 | |
| DIRECTDetector=Gemma 3, Future context=true2025.07 | 87.8 | |
| DIRECTDetector=Gemma 3, Future context=false2025.07 | 86.3 | |
| SCDetector=Qwen 2.5, Future context=false2025.07 | 85.9 |