Fact QA (Open Answer) on FAITHQA (Omission)
57Perfect ScoreLLAMA3-70B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| LLAMA3-70B2025.06 | 57 | 8.93 | |
| GPT-4o2025.06 | 49 | 8.62 | |
| LLAMA3-8B2025.06 | 46 | 8.52 | |
| Claude-32025.06 | 44 | 8.14 | |
| Claude-3.52025.06 | 37 | 6.73 | |
| GPT-4o-mini2025.06 | 36 | 7.86 | |
| Mistral-7B2025.06 | 20 | 7.15 |