Hallucination Detection on Earth Observation
90.94F1 ScoreGPT-5 nano
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GPT-5 nanoSize (B)=20*2026.03 | 90.94 | 5.33 | |
| GPT OSSSize (B)=120A52026.03 | 89.92 | 4.83 | |
| EVE-InstructSize (B)=242026.03 | 84.7 | 3.5 | |
| Qwen3Size (B)=235-A222026.03 | 84.4 | 2.17 | |
| MiniMax m2.5Size (B)=230A102026.03 | 83.77 | 5.17 | |
| GPT-4.1Size (B)=1800*2026.03 | 81.58 | 2.83 | |
| Mistral Medium 3.1Size (B)=200*2026.03 | 76.89 | 4.17 |