Binary Classification (Human vs Assistive) on All Datasets Combined
98AUCRole-Aware AI
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Role-Aware AIGenerator Model=LLaMA-3-Instruct, Generator Parameters=3 billion2026.05 | 98 | 94 | |
| RoBERTaGenerator Model=GPT-2, Generator Parameters=124 million2026.05 | 95 | 90 | |
| Role-Aware AIGenerator Model=GPT-2, Generator Parameters=124 million2026.05 | 93 | 88 | |
| LogLikelihoodGenerator Model=LLaMA-3-Instruct, Generator Parameters=3 billion2026.05 | 89 | 83 | |
| LogRankGenerator Model=LLaMA-3-Instruct, Generator Parameters=3 billion2026.05 | 88 | 81 | |
| LogRankGenerator Model=GPT-2, Generator Parameters=124 million2026.05 | 86 | 80 | |
| EntropyGenerator Model=LLaMA-3-Instruct, Generator Parameters=3 billion2026.05 | 84 | 78 | |
| CurvatureGenerator Model=LLaMA-3-Instruct, Generator Parameters=3 billion2026.05 | 75 | 70 | |
| RoBERTaGenerator Model=LLaMA-3-Instruct, Generator Parameters=3 billion2026.05 | 73 | 71 | |
| CurvatureGenerator Model=GPT-2, Generator Parameters=124 million2026.05 | 66 | 65 | |
| EntropyGenerator Model=GPT-2, Generator Parameters=124 million2026.05 | 65 | 64 | |
| LogLikelihoodGenerator Model=GPT-2, Generator Parameters=124 million2026.05 | 43 | 57 |