Multi-hop Reasoning on CRAG
30.08F1 ScoreTOTAL
Evaluation Results
| Method | Links | |
|---|---|---|
| TOTALModel=Claude, Prompting Method=TOTAL2025.10 | 30.08 | |
| TOTALModel=Gemini, Prompting Method=TOTAL2025.10 | 27.71 | |
| TOTALModel=GPT, Prompting Method=TOTAL2025.10 | 26.31 | |
| CIC + COTModel=Gemini, Prompting Method=CIC + COT2025.10 | 25.77 | |
| NAÏVEModel=GPT, Prompting Method=NAÏVE2025.10 | 25.73 | |
| CICModel=Gemini, Prompting Method=CIC2025.10 | 25.45 | |
| COTModel=Gemini, Prompting Method=COT2025.10 | 24.62 | |
| COTModel=GPT, Prompting Method=COT2025.10 | 23.24 | |
| CICModel=GPT, Prompting Method=CIC2025.10 | 22.12 | |
| NAÏVEModel=Gemini, Prompting Method=NAÏVE2025.10 | 22.03 | |
| CIC + COTModel=GPT, Prompting Method=CIC + COT2025.10 | 21.72 | |
| NAÏVEModel=Claude, Prompting Method=NAÏVE2025.10 | 20.49 | |
| COTModel=Claude, Prompting Method=COT2025.10 | 20.32 | |
| CIC + COTModel=Claude, Prompting Method=CIC + COT2025.10 | 18.86 | |
| CICModel=Claude, Prompting Method=CIC2025.10 | 17.32 |