Clinical Structured Data Extraction on synthetic
15TP CountInfherno
Evaluation Results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| InfhernoLLM=DeepSeek V3.1 Chat2025.07 | 15 | 1 | 1 | 93.8 | 93.8 | 93.8 | 0 | 100 | 96.8 | 16 | 16 | 16 | |
| InfhernoLLM=GPT-52025.07 | 15 | 1 | 1 | 93.8 | 93.8 | 93.8 | 0 | 100 | 96.8 | 16 | 12 | 9 | |
| InfhernoLLM=Qwen3-235B-A22B-25072025.07 | 15 | 1 | 1 | 93.8 | 93.8 | 93.8 | 0 | 100 | 96.8 | 16 | 15 | 15 | |
| InfhernoLLM=Qwen3-8B2025.07 | 15 | 1 | 1 | 93.8 | 93.8 | 93.8 | 0 | 100 | 96.8 | 16 | 15 | 15 | |
| InfhernoLLM=Gemini-2.5 Pro2025.07 | 14 | 1 | 1 | 93.3 | 93.3 | 93.3 | 0 | 100 | 96.6 | 15 | 15 | 15 | |
| InfhernoLLM=Claude Sonnet 4.52025.07 | 13 | 1 | 3 | 92.9 | 81.3 | 86.7 | 2 | 86.7 | 89.7 | 14 | 14 | 14 |