Red Flags on RealICU-GOLD
47.3HRR@5ICU-Evo
Evaluation Results
| Method | Links | |
|---|---|---|
| ICU-EvoBackbone=GPT-5.42026.05 | 47.3 | |
| ICU-EvoBackbone=Gemini-3.1-pro2026.05 | 30 | |
| Full-contextBackbone=GPT-5.42026.05 | 29.8 | |
| RAGBackbone=GPT-5.42026.05 | 23.4 | |
| RAGBackbone=Gemini-3.1-pro2026.05 | 21.6 | |
| Local-windowBackbone=GPT-5.42026.05 | 16.5 | |
| Local-windowBackbone=Gemini-3.1-pro2026.05 | 15.1 | |
| Full-contextBackbone=Gemini-3.1-pro2026.05 | 13.7 | |
| Full-contextBackbone=Qwen3-235B2026.05 | 11.7 | |
| ICU-EvoBackbone=Qwen3-235B2026.05 | 11.7 | |
| RAGBackbone=Qwen3-235B2026.05 | 9.5 | |
| Local-windowBackbone=Qwen3-235B2026.05 | 8 |