Long-form Biography Generation on Bio FactScore
81.2FactScoreSELF-RAG
Evaluation Results
| Method | Links | |
|---|---|---|
| SELF-RAGScale=7B, Retrieval=Yes2023.10 | 81.2 | |
| SELF-RAGScale=13B, Retrieval=Yes2023.10 | 80.2 | |
| Ret-Llama2-ChatScale=13B, Retrieval=Yes, Proprietary=true2023.10 | 79.9 | |
| Llama2-FTScale=7B, Retrieval=Yes, Fine-tuned=true2023.10 | 78.2 | |
| Llama2Scale=7B, Retrieval=Yes2023.10 | 78 | |
| AlpacaScale=13B, Retrieval=Yes2023.10 | 77.7 | |
| Llama2Scale=13B, Retrieval=Yes2023.10 | 77.5 | |
| AlpacaScale=7B, Retrieval=Yes2023.10 | 76.6 | |
| Ret-ChatGPTRetrieval=Yes, Proprietary=true2023.10 | 75.3 | |
| ChatGPTRetrieval=No, Proprietary=true2023.10 | 71.8 | |
| Perplexity.aiProprietary=true2023.10 | 71.2 | |
| CoVEScale=65B, Mode=Iterative Prompting2023.10 | 71.2 | |
| Llama2-ChatScale=13B, Retrieval=No, Proprietary=true2023.10 | 55.9 | |
| Llama2Scale=13B, Retrieval=No2023.10 | 53.4 | |
| AlpacaScale=13B, Retrieval=No2023.10 | 50.2 | |
| AlpacaScale=7B, Retrieval=No2023.10 | 45.8 | |
| Llama2Scale=7B, Retrieval=No2023.10 | 44.5 |