Fact-Checking on AVERICITEC Retrieved Evidence (test)
77.7AccuracyRADARmulti
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| RADARmultiBackbone=GPT-4o mini2026.04 | 77.7 | 63.3 | 45.6 | 56.5 | 87.6 | |
| RADARmultiBackbone=LLaMA3-8B-Instruct2026.04 | 77.3 | 63.2 | 46.3 | 56.7 | 86.5 | |
| RADARmultiBackbone=Qwen2.5-7B-Instruct2026.04 | 76.7 | 63.1 | 48.2 | 54.9 | 86.2 | |
| RADARsingleBackbone=Qwen2.5-7B-Instruct2026.04 | 71.2 | 54.9 | 39.4 | 41.1 | 84.3 | |
| D2DBackbone=Qwen2.5-7B-Instruct2026.04 | 69 | 48.5 | 22.2 | 41.2 | 82.1 | |
| D2DBackbone=LLaMA3-8B-Instruct2026.04 | 64.1 | 45.1 | 21.7 | 34.1 | 79.5 | |
| FIREBackbone=Qwen2.5-7B-Instruct2026.04 | 63.9 | 46.9 | 29.2 | 32.1 | 79.5 | |
| D2DBackbone=GPT-4o mini2026.04 | 63 | 50.9 | 36.3 | 39.7 | 76.5 | |
| FIREBackbone=GPT-4o mini2026.04 | 60.3 | 46.9 | 30.7 | 34.1 | 75.9 | |
| RADARsingleBackbone=GPT-4o mini2026.04 | 58.4 | 51 | 41.2 | 41.5 | 70.2 | |
| RADARsingleBackbone=LLaMA3-8B-Instruct2026.04 | 57.3 | 52 | 44.4 | 44.8 | 66.6 | |
| FIREBackbone=LLaMA3-8B-Instruct2026.04 | 50.4 | 43.4 | 30.9 | 39.7 | 59.6 |