Soft prediction on FABLES+
88AUROCVeriTrail (t = 2)
Evaluation results
| Method | Links | |
|---|---|---|
| VeriTrail (t = 2)threshold t=22025.05 | 88 | |
| VeriTrail (t = 1)threshold t=12025.05 | 86 | |
| VeriTrail (t = 3)threshold t=32025.05 | 85 | |
| RAGk=202025.05 | 81 | |
| Bespoke-MiniCheck-7Bk=4, Base Model=Llama-3.12025.05 | 79 | |
| GPT-4.1 Mini2025.05 | 67 | |
| INFUSE2025.05 | 64 | |
| AlignScorek=12025.05 | 57 |