AI-generated text detection on RealDet
95.32AUROCLAPD
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LAPDStandardization=true, Base Model=Llama2-7B, Aligned Model=Llama2-7B-Instruct2026.04 | 95.32 | — | |
| LAPD2026.04 | 95.32 | — | |
| Exons-DetectTraining paradigm=Training-free2026.03 | 94.98 | 91.3 | |
| DNA-DetectLLMTraining paradigm=Training-free2026.03 | 94.48 | 90.58 | |
| DNA-DetectLLMStandardization=true2026.04 | 94.48 | — | |
| Lastde++Training paradigm=Training-free2026.03 | 93.9 | 89.41 | |
| BinocularsStandardization=false2026.04 | 93.67 | — | |
| BinocularsTraining paradigm=Training-free2026.03 | 93.64 | 90.51 | |
| Fast-DetectGPTTraining paradigm=Training-free2026.03 | 93.25 | 90 | |
| BiscopeTraining paradigm=Training-based2026.03 | 92.88 | 86.9 | |
| Fast-DetectGPTStandardization=true2026.04 | 92.66 | — | |
| Lastde++Standardization=true2026.04 | 92.39 | — | |
| ReMoDetect2026.04 | 92.18 | — | |
| ImBD2026.04 | 92.12 | — | |
| EntropyStandardization=false2026.04 | 90.08 | — | |
| LogRankStandardization=false2026.04 | 89.93 | — | |
| LikelihoodStandardization=false2026.04 | 89.17 | — | |
| RAIStandardization=false, Base Model=Llama2-7B, Aligned Model=Llama2-7B-Instruct2026.04 | 86.92 | — | |
| LogRankTraining paradigm=Training-free2026.03 | 86.28 | 80.69 | |
| LikelihoodTraining paradigm=Training-free2026.03 | 85.35 | 79.75 | |
| OpenAI Detector2026.04 | 84.96 | — | |
| OpenAI-DTraining paradigm=Training-based2026.03 | 84.75 | 77.47 | |
| DetectGPTTraining paradigm=Training-free2026.03 | 78.69 | 73.8 | |
| IRMTraining paradigm=Training-free2026.03 | 77.7 | 76.62 | |
| EntropyTraining paradigm=Training-free2026.03 | 75.42 | 74.72 | |
| R-DetectTraining paradigm=Training-based2026.03 | 65.93 | 67.72 | |
| DetectGPTStandardization=true2026.04 | 61.33 | — |