Multi-hop Reasoning on FanOutQA
71.84F1 ScoreTOTAL
Evaluation Results
| Method | Links | |
|---|---|---|
| TOTALModel=Gemini, Prompting Method=TOTAL2025.10 | 71.84 | |
| TOTALModel=Claude, Prompting Method=TOTAL2025.10 | 69.99 | |
| TOTALModel=GPT, Prompting Method=TOTAL2025.10 | 69.07 | |
| CIC + COTModel=Gemini, Prompting Method=CIC + COT2025.10 | 66.97 | |
| CICModel=Gemini, Prompting Method=CIC2025.10 | 66.44 | |
| CIC + COTModel=GPT, Prompting Method=CIC + COT2025.10 | 66.35 | |
| CIC + COTModel=Claude, Prompting Method=CIC + COT2025.10 | 66.29 | |
| CICModel=Claude, Prompting Method=CIC2025.10 | 63.74 | |
| CICModel=GPT, Prompting Method=CIC2025.10 | 63.39 | |
| COTModel=GPT, Prompting Method=COT2025.10 | 49.09 | |
| NAÏVEModel=GPT, Prompting Method=NAÏVE2025.10 | 48.77 | |
| NAÏVEModel=Claude, Prompting Method=NAÏVE2025.10 | 46.72 | |
| NAÏVEModel=Gemini, Prompting Method=NAÏVE2025.10 | 46.54 | |
| COTModel=Claude, Prompting Method=COT2025.10 | 45.54 | |
| COTModel=Gemini, Prompting Method=COT2025.10 | 43.54 |