Soft prediction on DiverseSumm+
87AUROCVeriTrail (t = 2)
Evaluation results
| Method | Links | |
|---|---|---|
| VeriTrail (t = 2)threshold t=22025.05 | 87 | |
| VeriTrail (t = 3)threshold t=32025.05 | 80 | |
| VeriTrail (t = 1)threshold t=12025.05 | 79 | |
| Bespoke-MiniCheck-7Bk=1, Base Model=Llama-3.12025.05 | 79 | |
| RAGk=152025.05 | 76 | |
| AlignScorek=22025.05 | 68 | |
| GPT-4.1 Mini2025.05 | 61 | |
| INFUSE2025.05 | 58 |