| Benchmark | Model | Rank | Results |
|---|---|---|---|
| common-sense-reasoning-on-commonsenseqa | GrapeQA: PEGA | – | Accuracy: 73.5 |
| question-answering-on-medqa-usmle | GrapeQA: PEGA | – | Accuracy: 39.51 |
| question-answering-on-openbookqa | GrapeQA: CANP | – | Accuracy: 66.2 |
| question-answering-on-openbookqa | GrapeQA: PEGA | – | Accuracy: 82 |
| question-answering-on-openbookqa | GrapeQA: PEGA+CANP | – | Accuracy: 90 |