Diagnosis on Symptom2Disease (test)
98Diagnosis AccuracyCLIN-LLM
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CLIN-LLMUncertainty Estimation=Monte Carlo Dropout, Explainability=High (RAG + Summary)2025.10 | 98 | 98 | |
| Med-PaLMUncertainty Estimation=Some (self-consistency), Explainability=Limited auditability2025.10 | 96.1 | 95.1 | |
| GPT-5 (API)Uncertainty Estimation=Heuristic prompts, Explainability=Variable2025.10 | 94.1 | 93.3 | |
| BioClinicalBERTExplainability=Moderate (token-level)2025.10 | 93.2 | 92.1 | |
| ClinicalBERT2025.10 | 91 | 90.3 |