Factuality Hallucination on BioGEN
49.1FactScoreNOVA
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| NOVABackbone=Qwen-2, Selection Ratio=5%2025.02 | 49.1 | 100 | |
| NOVABackbone=Qwen-2, Selection Ratio=10%2025.02 | 47.5 | 100 | |
| NOVABackbone=Qwen-2, Selection Ratio=15%2025.02 | 47.2 | 100 | |
| FLAME-DPOBackbone=Qwen-2, Selection Ratio=100%2025.02 | 47.1 | 100 | |
| SELF-EVALBackbone=Qwen-2, Selection Ratio=100%2025.02 | 46.8 | 100 | |
| NuggetsBackbone=Qwen-2, Selection Ratio=5%2025.02 | 46.6 | 100 | |
| NuggetsBackbone=Qwen-2, Selection Ratio=10%2025.02 | 46.1 | 100 | |
| CaRBackbone=Qwen-2, Selection Ratio=5%2025.02 | 45.7 | 100 | |
| NuggetsBackbone=Qwen-2, Selection Ratio=15%2025.02 | 45.7 | 100 | |
| CaRBackbone=Qwen-2, Selection Ratio=10%2025.02 | 45.2 | 100 | |
| CaRBackbone=Qwen-2, Selection Ratio=15%2025.02 | 44.8 | 100 | |
| IFDBackbone=Qwen-2, Selection Ratio=10%2025.02 | 44.5 | 100 | |
| IFDBackbone=Qwen-2, Selection Ratio=5%2025.02 | 44.2 | 100 | |
| NOVABackbone=LLaMA-1, Selection Ratio=15%2025.02 | 44.1 | 100 | |
| IFDBackbone=Qwen-2, Selection Ratio=15%2025.02 | 43.7 | 100 | |
| NOVABackbone=LLaMA-1, Selection Ratio=5%2025.02 | 43.6 | 100 | |
| NOVABackbone=LLaMA-1, Selection Ratio=10%2025.02 | 43.2 | 100 | |
| SELF-EVALBackbone=LLaMA-1, Selection Ratio=100%2025.02 | 41.8 | 100 | |
| FLAME-DPOBackbone=LLaMA-1, Selection Ratio=100%2025.02 | 41.2 | 100 | |
| NuggetsBackbone=LLaMA-1, Selection Ratio=10%2025.02 | 41 | 100 | |
| NuggetsBackbone=LLaMA-1, Selection Ratio=15%2025.02 | 40.9 | 100 | |
| IFDBackbone=LLaMA-1, Selection Ratio=10%2025.02 | 40.7 | 100 | |
| CaRBackbone=LLaMA-1, Selection Ratio=10%2025.02 | 40.3 | 100 | |
| VanillaBackbone=Qwen-2, Selection Ratio=100%2025.02 | 40.3 | 100 | |
| IFDBackbone=LLaMA-1, Selection Ratio=5%2025.02 | 40.2 | 100 | |
| CaRBackbone=LLaMA-1, Selection Ratio=15%2025.02 | 40.2 | 100 | |
| CaRBackbone=LLaMA-1, Selection Ratio=5%2025.02 | 39.6 | 100 | |
| NuggetsBackbone=LLaMA-1, Selection Ratio=5%2025.02 | 39.3 | 100 | |
| IFDBackbone=LLaMA-1, Selection Ratio=15%2025.02 | 39.2 | 100 | |
| VanillaBackbone=LLaMA-1, Selection Ratio=100%2025.02 | 38.6 | 100 |