Rare disease mention extraction on MIMIC3 RD Code n=79 docs (test)
52.6Micro-F1 ScoreMistral 24B
Evaluation Results
| Method | Links | |
|---|---|---|
| Mistral 24BApproach=RDMA2025.07 | 52.6 | |
| Qwen3 32BApproach=RDMA2025.07 | 46.7 | |
| GPT-5Approach=RDMA2025.07 | 46.5 | |
| Llama-3 70BApproach=RDMA2025.07 | 43.9 | |
| GPT-5Approach=RAG2025.07 | 39.1 | |
| DictionaryApproach=Baselines2025.07 | 38.4 | |
| Qwen3 32BApproach=RAG2025.07 | 14.3 | |
| Llama-3 70BApproach=RAG2025.07 | 13.9 | |
| GPT-5Approach=Zeroshot2025.07 | 9.8 | |
| Llama-3 8BApproach=RDMA2025.07 | 9.7 | |
| BioClinicalBERTApproach=Baselines2025.07 | 5 | |
| BioBERT-MRCApproach=Baselines2025.07 | 4.6 | |
| Mistral 24BApproach=RAG2025.07 | 3.2 | |
| Llama-3 8BApproach=RAG2025.07 | 2.5 | |
| i2b2Approach=Baselines2025.07 | 1.9 | |
| Mistral 24BApproach=Zeroshot2025.07 | 0.2 | |
| Llama-3 70BApproach=Zeroshot2025.07 | 0 | |
| Llama-3 8BApproach=Zeroshot2025.07 | 0 | |
| Qwen3 32BApproach=Zeroshot2025.07 | 0 |