Meteorological diagnosis agentic workflow on HVR-Met (100 extreme weather cases)
4.7Hypothesis ScoreGPT-5
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| GPT-5Model=GPT-52026.03 | 4.7 | 4.92 | 4.07 | 4.13 | 3.94 | |
| GEMINI-3-PRO-PROVIEW-THINKINGModel=GEMINI-3-PRO-PROVIEW-THINKING2026.03 | 4.63 | 4.87 | 4.02 | 3.98 | 3.76 | |
| DEEPSEEK-R1-0528Model=DEEPSEEK-R1-05282026.03 | 3.91 | 4.25 | 3.56 | 3.41 | 2.97 | |
| QWEN3-CODER-480B-A35B-INSTRUCTModel=QWEN3-CODER-480B-A35B-INSTRUCT2026.03 | 3.73 | 4.31 | 3.3 | 2.46 | 2.33 |