Question Answering on AstroQA Easy
77.5AccuracyMistral-7B + AstroRAG
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Mistral-7B + AstroRAGBackbone=ViT, Zero-shot=true2026.05 | 77.5 | 77.45 | |
| AstroRAGBackbone=Mistral-7B2026.05 | 77.5 | 77.45 | |
| AstroSage-Llama-3.1-8B + AstroRAGBackbone=MPT, Zero-shot=true2026.05 | 76.03 | 76.15 | |
| Map-Reduce2026.05 | 64.56 | 64.94 | |
| DrQA2026.05 | 62.06 | 62.43 | |
| Llama 2-Chat + AstroRAGBackbone=MPT, Zero-shot=true2026.05 | 52.94 | 53.63 | |
| AstroSage-Llama-3.1-8BBackbone=MPT, Zero-shot=true2026.05 | 41.47 | 41.08 | |
| Llama 2-ChatBackbone=MPT, Zero-shot=true2026.05 | 28.24 | 24.06 | |
| Mistral-7BBackbone=ViT, Zero-shot=true2026.05 | 20.74 | 8.6 | |
| BaselineBackbone=Mistral-7B2026.05 | 20.74 | 8.6 |