Sentence-level AI-generated text detection on MOSAIC XD
94.8Macro-F1SenFlow
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SenFlowType=TB, Evaluation Protocol=Unified2026.06 | 94.8 | 98.3 | 0.895 | |
| SenDetEXType=TB, Evaluation Protocol=Unified2026.06 | 93.3 | 98.1 | 0.867 | |
| SeqXGPTType=TB, Evaluation Protocol=Unified2026.06 | 84.1 | 93.2 | 0.687 | |
| POGERType=TB, Evaluation Protocol=Unified2026.06 | 80 | 89 | 0.624 | |
| BinocularsType=TF, Evaluation Protocol=Unified, Threshold tuning strategy=single threshold tuned on the combined validation set2026.06 | 62.1 | 68.3 | 0.243 | |
| Fast-DetectGPTType=TF, Evaluation Protocol=Unified, Threshold tuning strategy=single threshold tuned on the combined validation set2026.06 | 55.8 | 54.6 | 0.116 |