Machine-Generated Text Detection on MAGE eli5 (held-out)
0.958AUROCconfidence-weighted ensemble
Evaluation results
| Method | Links | ||
|---|---|---|---|
| confidence-weighted ensembleEnsemble Strategy=Confidence-weighted (Eq. 1)2026.07 | 0.958 | 63.6 | |
| Vanilla RoBERTa baselineBackbone=RoBERTa-large2026.07 | 0.957 | 66.2 | |
| FOMAML+LoRAK-shot=102026.07 | 0.936 | 48.4 |