Verification on SummEval (AUROC)
0.755AUROCVERDI LR
Evaluation Results
| Method | Links | |
|---|---|---|
| VERDI LRModel=GPT-5.4-mini2026.05 | 0.755 | |
| VERDI LRModel=GPT-4.1-mini2026.05 | 0.717 | |
| VERDI LRModel=Qwen3.5-4B2026.05 | 0.681 | |
| VERDI LRModel=Qwen3.5-9B2026.05 | 0.654 | |
| VERDI LRModel=Qwen3.5-27B2026.05 | 0.637 | |
| LogprobModel=Qwen3.5-9B2026.05 | 0.417 | |
| LogprobModel=Qwen3.5-4B2026.05 | 0.373 | |
| LogprobModel=Qwen3.5-27B2026.05 | 0.325 |