Data Contamination Detection on Seen vs. Unseen Datasets (dataset-level)
100AUCCoDeC
Evaluation Results
| Method | Links | |
|---|---|---|
| CoDeCModel=Pythia 410M2025.10 | 100 | |
| CoDeCModel=Pythia 1.4B2025.10 | 100 | |
| CoDeCModel=Pythia 12B2025.10 | 100 | |
| CoDeCModel=GPT-Neo 1.3B2025.10 | 100 | |
| CoDeCModel=GPT-Neo 20B2025.10 | 100 | |
| CoDeCModel=RWKV-4 430M2025.10 | 100 | |
| CoDeCModel=RWKV-4 3B2025.10 | 100 | |
| CoDeCModel=OLMo 1B2025.10 | 100 | |
| CoDeCModel=OLMo 7B2025.10 | 100 | |
| CoDeCModel=Nemotron v2 9B2025.10 | 100 | |
| CoDeCModel=Nemotron-H 8B2025.10 | 100 | |
| CoDeCModel=Nemotron-H 56B2025.10 | 100 | |
| CoDeCModel=Cumulative2025.10 | 99.9 | |
| CoDeCModel=RWKV-4 14B2025.10 | 99.6 | |
| ZlibModel=Nemotron v2 9B2025.10 | 98.2 | |
| Min-K%Model=Nemotron v2 9B2025.10 | 97.6 | |
| ZlibModel=GPT-Neo 20B2025.10 | 92.7 | |
| ZlibModel=RWKV-4 14B2025.10 | 92.7 | |
| ZlibModel=Pythia 410M2025.10 | 92.3 | |
| ZlibModel=Pythia 12B2025.10 | 92.3 | |
| ZlibModel=RWKV-4 430M2025.10 | 92.3 | |
| ZlibModel=Nemotron-H 56B2025.10 | 92 | |
| ZlibModel=RWKV-4 3B2025.10 | 91.9 | |
| ZlibModel=Pythia 1.4B2025.10 | 91.5 | |
| ZlibModel=GPT-Neo 1.3B2025.10 | 90.8 | |
| ZlibModel=Cumulative2025.10 | 89.6 | |
| ZlibModel=Nemotron-H 8B2025.10 | 88 | |
| Min-K%Model=Nemotron-H 56B2025.10 | 86.7 | |
| Vanilla lossModel=Nemotron v2 9B2025.10 | 83.7 | |
| Min-K%Model=GPT-Neo 20B2025.10 | 83.5 | |
| Min-K%Model=Pythia 12B2025.10 | 82.3 | |
| Vanilla lossModel=Nemotron-H 56B2025.10 | 82.2 | |
| Min-K%Model=RWKV-4 14B2025.10 | 81.5 | |
| ZlibModel=OLMo 1B2025.10 | 80.5 | |
| Vanilla lossModel=Nemotron-H 8B2025.10 | 80 | |
| Min-K%Model=RWKV-4 3B2025.10 | 79.6 | |
| Min-K%Model=Pythia 1.4B2025.10 | 79.2 | |
| Min-K%Model=GPT-Neo 1.3B2025.10 | 79.2 | |
| Min-K%Model=Cumulative2025.10 | 78.5 | |
| ZlibModel=OLMo 7B2025.10 | 78.1 | |
| Vanilla lossModel=Pythia 1.4B2025.10 | 77.3 | |
| Vanilla lossModel=GPT-Neo 1.3B2025.10 | 77.3 | |
| Vanilla lossModel=RWKV-4 14B2025.10 | 77.3 | |
| Vanilla lossModel=Pythia 12B2025.10 | 76.9 | |
| Vanilla lossModel=GPT-Neo 20B2025.10 | 76.9 | |
| Vanilla lossModel=RWKV-4 3B2025.10 | 76.5 | |
| Min-K%Model=Pythia 410M2025.10 | 76.2 | |
| Vanilla lossModel=Cumulative2025.10 | 75.7 | |
| Vanilla lossModel=RWKV-4 430M2025.10 | 75.4 | |
| Min-K%Model=RWKV-4 430M2025.10 | 75.4 | |
| Vanilla lossModel=Pythia 410M2025.10 | 75 | |
| Min-K%Model=Nemotron-H 8B2025.10 | 73.3 | |
| Min-K%Model=OLMo 7B2025.10 | 72.7 | |
| Min-K%Model=OLMo 1B2025.10 | 71.9 | |
| Vanilla lossModel=OLMo 7B2025.10 | 65.6 | |
| Vanilla lossModel=OLMo 1B2025.10 | 64.8 |