AI-Generated Text Detection on SeqXGPT-bench (test)
95.99Precision (AI)Deberta
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Deberta2024.06 | 95.99 | 95.83 | 95.84 | 96 | 95.92 | |
| SCRN2024.06 | 95.69 | 96.17 | 96.15 | 95.67 | 95.92 | |
| SeqXGPT2024.06 | 95.03 | 97.39 | 97.32 | 94.93 | 96.15 | |
| Flooding2024.06 | 94.91 | 96.33 | 96.28 | 94.83 | 95.58 | |
| Roberta2024.06 | 94.77 | 96.66 | 96.6 | 94.67 | 95.67 | |
| RDrop2024.06 | 94.63 | 97 | 96.92 | 94.5 | 95.75 | |
| Bert2024.06 | 86.76 | 90.67 | 90.23 | 86.17 | 88.5 | |
| RMLM2024.06 | 85.34 | 92.17 | 91.49 | 84.17 | 88.15 | |
| RanMASK2024.06 | 78.11 | 95.17 | 93.82 | 73.33 | 84.06 | |
| GLTR2024.06 | 69.01 | 73.5 | 71.66 | 67 | 70.22 | |
| Log-Rank2024.06 | 65.69 | 70.83 | 68.35 | 63 | 66.87 | |
| Log-Likelihood2024.06 | 64.89 | 69 | 66.9 | 62.67 | 65.8 | |
| Entropy2024.06 | 57.26 | 59.17 | 57.76 | 55.83 | 57.49 |