Machine-Generated Text Detection on Essay ChatGLM
98.08TPR@FPR=1%Likelihood-M
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Likelihood-MTraining Source=Claude, Detection Method Variant=Markov-informed score calibration2026.02 | 98.08 | 99.63 | |
| Log-RankTraining Source=Claude, Detection Method Variant=Standard2026.02 | 95.71 | 99.36 | |
| Log-Rank-MTraining Source=Claude, Detection Method Variant=Markov-informed score calibration2026.02 | 95 | 99.25 | |
| LikelihoodTraining Source=Claude, Detection Method Variant=Standard2026.02 | 92.86 | 99.29 | |
| DetectGPT-MTraining Source=Claude, Detection Method Variant=Markov-informed score calibration2026.02 | 74.06 | 96.46 | |
| FastGPT-MTraining Source=Claude, Detection Method Variant=Markov-informed score calibration2026.02 | 56.47 | 95.35 | |
| Entropy-MTraining Source=Claude, Detection Method Variant=Markov-informed score calibration2026.02 | 30.94 | 90.88 | |
| DNA-GPT-MTraining Source=Claude, Detection Method Variant=Markov-informed score calibration2026.02 | 12.14 | 75.98 | |
| EntropyTraining Source=Claude, Detection Method Variant=Standard2026.02 | 6.88 | 84.11 | |
| FastGPTTraining Source=Claude, Detection Method Variant=Standard2026.02 | 2.72 | 71.08 | |
| DetectGPTTraining Source=Claude, Detection Method Variant=Standard2026.02 | 0 | 49.25 | |
| DNA-GPTTraining Source=Claude, Detection Method Variant=Standard2026.02 | 0 | 36.19 |