Fact-Checking on COVID-Fact
75.9Balanced AccOpenAI o1
Evaluation results
| Method | Links | |
|---|---|---|
| OpenAI o1Scale=>300B2025.02 | 75.9 | |
| DeepSeek-V3 671BScale=>300B2025.02 | 75.9 | |
| CoTBackbone=DeepSeek-V3-671B, Direct Decision=true2026.05 | 75.4 | |
| VanillaBackbone=DeepSeek-V3-671B, Direct Decision=false2026.05 | 75.2 | |
| GPT-4Scale=>300B2025.02 | 73.8 | |
| Claude 3.5-SonnetScale=>300B2025.02 | 73.8 | |
| ANCHORBackbone=Qwen2.5-72B, Direct Decision=false2026.05 | 72.6 | |
| CoTBackbone=Qwen2.5-72B, Direct Decision=false2026.05 | 72.1 | |
| GraphCheck-Llama3.3 70BBackbone=Llama3.3 70B2025.02 | 71.5 | |
| GraphCheckBackbone=Llama3.3 70B2025.02 | 71.5 | |
| CoTBackbone=Qwen2.5-72B, Direct Decision=true2026.05 | 70.1 | |
| Qwen2.5 72BBackbone=Qwen2.5 72B2025.02 | 69.9 | |
| Qwen2.5 72B2025.02 | 69.9 | |
| GraphCheck-Qwen2.5 72BBackbone=Qwen2.5 72B2025.02 | 69.7 | |
| GraphCheckBackbone=Qwen 72B2025.02 | 69.7 | |
| Llama3.3 70BBackbone=Llama3.3 70B2025.02 | 69 | |
| Llama3.3 70B2025.02 | 69 | |
| VanillaBackbone=DeepSeek-V3-671B, Direct Decision=true2026.05 | 68.6 | |
| VanillaBackbone=Qwen2.5-72B, Direct Decision=true2026.05 | 68.6 | |
| CoTBackbone=DeepSeek-V3-671B, Direct Decision=false2026.05 | 68.5 | |
| AlignScore2025.02 | 66.5 | |
| AlignScore2025.02 | 66.5 | |
| BIRDBackbone=Qwen2.5-72B, Direct Decision=false2026.05 | 66.4 | |
| MiniCheck2025.02 | 65.9 | |
| MiniCheck2025.02 | 65.9 | |
| ACUEval2025.02 | 64.7 | |
| GPT-4oScale=>300B2025.02 | 62.6 | |
| GraphEval2025.02 | 60.7 | |
| GraphEval2025.02 | 60.7 | |
| Qwen2.5 7B2025.02 | 59.2 | |
| Llama3 8B2025.02 | 58.1 | |
| VanillaBackbone=Qwen2.5-72B, Direct Decision=false2026.05 | 53.8 |