Scientific Reasoning on GPQA (Overall)
51.01Overall ScoreRAISE
Evaluation Results
| Method | Links | |
|---|---|---|
| RAISEStrategy Category=Ours2025.06 | 51.01 | |
| HyDEStrategy Category=Decomposed+RAG2025.06 | 46.46 | |
| CoT+RAGStrategy Category=Direct+RAG2025.06 | 45.96 | |
| Least-to-Most+RAGStrategy Category=Decomposed+RAG2025.06 | 45.95 | |
| Least-to-MostStrategy Category=Decomposed2025.06 | 44.95 | |
| Step-BackStrategy Category=Decomposed2025.06 | 44.44 | |
| Step-Back+RAGStrategy Category=Decomposed+RAG2025.06 | 43.43 | |
| CoTStrategy Category=Direct2025.06 | 42.42 |