Question Answering on ResearchQA (RQA) Artificial Intelligence (test)
79.3Rubrics Scoreo3 + intent
Evaluation results
| Method | Links | |
|---|---|---|
| o3 + intentintent-aware framework=true2026.03 | 79.3 | |
| o3intent-aware framework=false2026.03 | 76.3 | |
| Claude opus-4 + intentintent-aware framework=true2026.03 | 75.7 | |
| Claude opus-4intent-aware framework=false2026.03 | 74.3 | |
| gemini-2.5-pro + intentintent-aware framework=true2026.03 | 74 | |
| gemini-2.5-prointent-aware framework=false2026.03 | 71.9 |