Science Question Answering on AI2ARC (test)
Evaluation results
| Method | Links | |
|---|
| GPT-4number_of_shots=25-shot2023.03 | | 96.3 |
|---|
| ST-MoEbenchmark-specific tuning=true2023.03 | | 86.5 |
|---|
| GPT-3.5number_of_shots=25-shot2023.03 | | 85.2 |
|---|
| PaLMnumber_of_shots=8-shot2023.03 | | 85.2 |
|---|
| TAMA2025.01 | | 81.23 |
|---|
| base2025.01 | | 80.89 |
|---|