Machine-Generated Text Detection on M4 (ChatGPT)
0.8357AUROC (Arabic)VaryBalance
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| VaryBalance2026.01 | 0.8357 | 0.8821 | 0.9867 | 0.8581 | |
| Binoculars2026.01 | 0.8355 | 0.8658 | 0.942 | 0.7749 | |
| Fast-DetectGPT2026.01 | 0.6146 | 0.8692 | 0.7846 | 0.4416 | |
| AdaDetectGPT2026.01 | 0.5632 | 0.8847 | 0.6784 | 0.4728 | |
| RADAR2026.01 | 0.357 | 0.6969 | 0.4462 | 0.3631 | |
| DNA-GPT2026.01 | 0.2117 | 0.3145 | 0.4571 | 0.631 |