Veracity Assessment on FacTool-QA
92True F1MERMAID
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| MERMAIDLLM=GPT-4o2026.01 | 92 | 68 | 80 | |
| MERMAIDLLM=GPT-5-mini2026.01 | 90 | 66 | 78 | |
| FIRELLM=GPT-4o2026.01 | 89 | 66 | 77 | |
| SAFELLM=GPT-4o2026.01 | 88 | 63 | 76 | |
| MERMAIDLLM=OSS-120B2026.01 | 88 | 56 | 72 | |
| MERMAIDLLM=Qwen-2.5-70B2026.01 | 86 | 59 | 73 | |
| MERMAIDLLM=OSS-20B2026.01 | 86 | 54 | 70 | |
| MERMAIDLLM=Qwen-2.5-7B2026.01 | 85 | 54 | 70 | |
| FacToolLLM=GPT-4o2026.01 | 84 | 58 | 71 | |
| FactCheck-GPTLLM=GPT-4o2026.01 | 84 | 60 | 72 | |
| MERMAIDLLM=LLaMA-3.1-70B-Inst2026.01 | 69 | 44 | 57 | |
| MERMAIDLLM=LLaMA-3.1-8B-Inst2026.01 | 63 | 41 | 52 |