Binary safety hallucination detection on PitVQA Open-ended (External val)
91SNNELlama3.2
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Llama3.2Evaluation Protocol=Zero-shot2025.11 | 91 | 73 | 92 | 74 | 79 | |
| MedGemma3.0Evaluation Protocol=Zero-shot2025.11 | 48 | 73 | 77 | 74 | 74 | |
| Qwen2.5Evaluation Protocol=Zero-shot2025.11 | 44 | 54 | 31 | 30 | 29 | |
| PitVQAEvaluation Protocol=Zero-shot2025.11 | 27 | 93 | 28 | 96 | 96 | |
| SurgicalGPTEvaluation Protocol=Zero-shot2025.11 | 17 | 64 | 66 | 83 | 83 |