Fact-checking on AVeriTeC (dev)
0.46Hu-METEOR (Q only)GPT-4o
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GPT-4oPipeline configuration=full-featured2024.10 | 0.46 | 0.29 | 0.42 | |
| Llama 3.1 70BPipeline configuration=full-featured2024.10 | 0.46 | 0.27 | 0.36 | |
| GPT-4oPipeline configuration=simplified2024.10 | 0.45 | 0.28 | 0.38 | |
| GPT-4oPipeline configuration=with DeBERTa classification2024.10 | 0.45 | 0.28 | 0.36 | |
| Claude-3.5-SonnetPipeline configuration=full-featured2024.10 | 0.43 | 0.28 | 0.35 | |
| AVeriTeC baseline2024.10 | 0.24 | 0.19 | 0.09 |