Machine-generated text detection on GovernmentPublic Llama-3-70B-Instruct (test)
97.2AUCLearn-to-Distance
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Learn-to-DistanceEvaluation Protocol=Cross-fitting2026.01 | 97.2 | 1.057 | 27.3 | |
| ImBDEvaluation Protocol=Cross-fitting2026.01 | 96.1 | — | — | |
| FDGPTEvaluation Protocol=Cross-fitting2026.01 | 87.1 | — | — | |
| BinocularsEvaluation Protocol=Cross-fitting2026.01 | 87 | — | — | |
| LikelihoodEvaluation Protocol=Cross-fitting2026.01 | 73.1 | — | — | |
| BARTScoreEvaluation Protocol=Cross-fitting2026.01 | 71.8 | — | — | |
| LRREvaluation Protocol=Cross-fitting2026.01 | 71.2 | — | — | |
| RAIDAREvaluation Protocol=Cross-fitting2026.01 | 63.4 | — | — | |
| IDEEvaluation Protocol=Cross-fitting2026.01 | 61.5 | — | — | |
| RoBERTaEvaluation Protocol=Cross-fitting2026.01 | 57.2 | — | — | |
| RADAREvaluation Protocol=Cross-fitting2026.01 | 57.1 | — | — | |
| ADGPTEvaluation Protocol=Cross-fitting2026.01 | 56.4 | — | — |