Question Answering on HotpotQA (ASR, RSR, PPL)
90ASRAuthChain
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| AuthChainModel=GPT-3.52025.05 | 90 | — | — | |
| AuthChainModel=GPT-4o2025.05 | 88 | — | — | |
| AuthChainModel=Deepseek-V3-03242025.05 | 87 | — | — | |
| AuthChainModel=GPT-42025.05 | 86 | — | — | |
| AuthChainModel=Llama3-70B2025.05 | 86 | — | — | |
| AuthChainModel=Llama3-8B2025.05 | 85 | — | — | |
| AuthorityModel=GPT-3.52025.05 | 82 | — | — | |
| AuthorityModel=Llama3-70B2025.05 | 80 | — | — | |
| AuthorityModel=Deepseek-V3-03242025.05 | 80 | — | — | |
| CoEModel=GPT-3.52025.05 | 79 | — | — | |
| AuthorityModel=GPT-4o2025.05 | 78 | — | — | |
| AuthorityModel=Llama3-8B2025.05 | 78 | — | — | |
| HRAGModel=GPT-42025.05 | 77 | — | — | |
| CoEModel=Deepseek-V3-03242025.05 | 77 | — | — | |
| CoEModel=Llama3-70B2025.05 | 75 | — | — | |
| HRAGModel=Deepseek-V3-03242025.05 | 73 | — | — | |
| CoEModel=GPT-4o2025.05 | 71 | — | — | |
| CoEModel=Llama3-8B2025.05 | 71 | — | — | |
| PRAGModel=GPT-3.52025.05 | 69 | — | — | |
| AuthorityModel=GPT-42025.05 | 66 | — | — | |
| PRAGModel=Deepseek-V3-03242025.05 | 63 | — | — | |
| CoEModel=GPT-42025.05 | 62 | — | — | |
| PRAGModel=Llama3-8B2025.05 | 62 | — | — | |
| PRAGModel=Llama3-70B2025.05 | 60 | — | — | |
| HRAGModel=Llama3-8B2025.05 | 58 | — | — | |
| HRAGModel=GPT-3.52025.05 | 57 | — | — | |
| PRAGModel=GPT-4o2025.05 | 54 | — | — | |
| PRAGModel=GPT-42025.05 | 49 | — | — | |
| HRAGModel=GPT-4o2025.05 | 49 | — | — | |
| HRAGModel=Llama3-70B2025.05 | 22 | — | — | |
| AuthChainModel=-2025.05 | — | 98 | 31 | |
| AuthorityModel=-2025.05 | — | 99 | 89 | |
| CoEModel=-2025.05 | — | 99 | 28.1 | |
| HRAGModel=-2025.05 | — | 100 | 352 | |
| PRAGModel=-2025.05 | — | 100 | 52.3 |