LGT Detection on Fast-DetectGPT PubMed (test)
0.97AUROCReMoDetect
Evaluation Results
| Method | Links | |
|---|---|---|
| ReMoDetectSource LLM=GPT4 Turbo2024.05 | 0.97 | |
| ReMoDetectModel=GPT4 Turbo2024.05 | 0.97 | |
| ReMoDetectSource LLM=GPT 3.5 Turbo2024.05 | 0.964 | |
| ReMoDetectSource LLM=Claude3 Opus2024.05 | 0.964 | |
| ReMoDetectModel=GPT3.5 Turbo2024.05 | 0.964 | |
| ReMoDetectModel=Claude 3 Opus2024.05 | 0.964 | |
| ReMoDetectSource LLM=Llama3 70B2024.05 | 0.963 | |
| ReMoDetectModel=Llama3 70B2024.05 | 0.963 | |
| ReMoDetectSource LLM=GPT42024.05 | 0.961 | |
| ReMoDetectModel=GPT42024.05 | 0.961 | |
| ReMoDetectSource Model=Gemma2 9B-it2024.05 | 0.951 | |
| ReMoDetectSource Model=Llama3 8B-it2024.05 | 0.946 | |
| ReMoDetectSource Model=OLMo 7B-sft2024.05 | 0.941 | |
| Fast-DetectGPTSource Model=OLMo 7B-sft2024.05 | 0.928 | |
| ReMoDetectSource Model=Qwen2 1.5B-it2024.05 | 0.927 | |
| Fast-DetectGPTModel=Llama3 70B2024.05 | 0.908 | |
| Fast-DetectGPTModel=GPT3.5 Turbo2024.05 | 0.902 | |
| GPTZeroSource LLM=Llama3 70B2024.05 | 0.901 | |
| ReMoDetectSource Model=Gemma2 2B-it2024.05 | 0.9 | |
| Fast-DetectGPTSource Model=Qwen2 1.5B-it2024.05 | 0.898 | |
| Fast-DetectGPTSource Model=Llama3 8B-it2024.05 | 0.896 | |
| Fast-DetectGPTModel=GPT4 Turbo2024.05 | 0.888 | |
| Log-likelihoodSource Model=OLMo 7B-sft2024.05 | 0.884 | |
| Fast-DetectGPTModel=Claude 3 Opus2024.05 | 0.882 | |
| GPTZeroSource LLM=GPT 3.5 Turbo2024.05 | 0.88 | |
| GPTZeroSource LLM=Claude3 Opus2024.05 | 0.88 | |
| Log-likelihoodModel=GPT3.5 Turbo2024.05 | 0.878 | |
| GPTZeroSource LLM=GPT4 Turbo2024.05 | 0.872 | |
| Log-likelihoodModel=GPT4 Turbo2024.05 | 0.865 | |
| ReMoDetectSource LLM=Gemini pro2024.05 | 0.864 | |
| ReMoDetectModel=Gemini pro2024.05 | 0.864 | |
| Log-likelihoodModel=Claude 3 Opus2024.05 | 0.855 | |
| Log-likelihoodModel=Llama3 70B2024.05 | 0.854 | |
| ChatGPT-DetectorSource Model=Gemma2 9B-it2024.05 | 0.851 | |
| Log-likelihoodSource Model=Llama3 8B-it2024.05 | 0.85 | |
| Fast-DetectGPTModel=GPT42024.05 | 0.85 | |
| GPTZeroSource LLM=GPT42024.05 | 0.848 | |
| GPTZeroSource LLM=Gemini pro2024.05 | 0.832 | |
| Log-likelihoodModel=Gemini pro2024.05 | 0.83 | |
| Log-likelihoodSource Model=Qwen2 1.5B-it2024.05 | 0.823 | |
| Fast-DetectGPTModel=Gemini pro2024.05 | 0.821 | |
| Log-likelihoodModel=GPT42024.05 | 0.81 | |
| ChatGPT-DetectorSource Model=Gemma2 2B-it2024.05 | 0.781 | |
| LRRModel=Gemini pro2024.05 | 0.75 | |
| DetectGPTModel=GPT3.5 Turbo2024.05 | 0.744 | |
| LRRModel=GPT3.5 Turbo2024.05 | 0.743 | |
| LRRModel=Claude 3 Opus2024.05 | 0.743 | |
| LRRModel=GPT4 Turbo2024.05 | 0.735 | |
| Fast-DetectGPTSource Model=Gemma2 2B-it2024.05 | 0.723 | |
| Fast-DetectGPTSource Model=Gemma2 9B-it2024.05 | 0.716 | |
| LRRModel=Llama3 70B2024.05 | 0.713 | |
| Log-likelihoodSource Model=Gemma2 9B-it2024.05 | 0.698 | |
| DetectGPTModel=GPT42024.05 | 0.681 | |
| LRRModel=GPT42024.05 | 0.681 | |
| Log-likelihoodSource Model=Gemma2 2B-it2024.05 | 0.679 | |
| NPRModel=GPT3.5 Turbo2024.05 | 0.678 | |
| NPRModel=Gemini pro2024.05 | 0.668 | |
| DetectGPTModel=Claude 3 Opus2024.05 | 0.663 | |
| DetectGPTModel=Llama3 70B2024.05 | 0.66 | |
| NPRModel=Llama3 70B2024.05 | 0.65 | |
| NPRModel=Claude 3 Opus2024.05 | 0.644 | |
| NPRModel=GPT4 Turbo2024.05 | 0.637 | |
| DetectGPTModel=GPT4 Turbo2024.05 | 0.636 | |
| NPRModel=GPT42024.05 | 0.633 | |
| DetectGPTModel=Gemini pro2024.05 | 0.632 | |
| OpenAI-DetectorSource Model=Qwen2 1.5B-it2024.05 | 0.629 | |
| OpenAI-DetectorSource Model=OLMo 7B-sft2024.05 | 0.62 | |
| OpenAI-DetectorModel=GPT3.5 Turbo2024.05 | 0.619 | |
| RankSource Model=Qwen2 1.5B-it2024.05 | 0.61 | |
| RankModel=Llama3 70B2024.05 | 0.609 | |
| RankModel=GPT4 Turbo2024.05 | 0.608 | |
| RankSource Model=OLMo 7B-sft2024.05 | 0.605 | |
| RankSource Model=Llama3 8B-it2024.05 | 0.604 | |
| RankModel=Claude 3 Opus2024.05 | 0.603 | |
| RankModel=GPT3.5 Turbo2024.05 | 0.598 | |
| RankModel=GPT42024.05 | 0.597 | |
| RankModel=Gemini pro2024.05 | 0.583 | |
| OpenAI-DetectorModel=Gemini pro2024.05 | 0.573 | |
| RankSource Model=Gemma2 2B-it2024.05 | 0.566 | |
| RankSource Model=Gemma2 9B-it2024.05 | 0.559 | |
| OpenAI-DetectorModel=GPT4 Turbo2024.05 | 0.558 | |
| OpenAI-DetectorSource Model=Llama3 8B-it2024.05 | 0.537 | |
| OpenAI-DetectorModel=GPT42024.05 | 0.531 | |
| OpenAI-DetectorModel=Llama3 70B2024.05 | 0.529 | |
| OpenAI-DetectorModel=Claude 3 Opus2024.05 | 0.489 | |
| OpenAI-DetectorSource Model=Gemma2 2B-it2024.05 | 0.444 | |
| ChatGPT-DetectorModel=Gemini pro2024.05 | 0.393 | |
| OpenAI-DetectorSource Model=Gemma2 9B-it2024.05 | 0.364 | |
| ChatGPT-DetectorModel=Llama3 70B2024.05 | 0.351 | |
| ChatGPT-DetectorSource Model=Llama3 8B-it2024.05 | 0.334 | |
| ChatGPT-DetectorModel=Claude 3 Opus2024.05 | 0.331 | |
| ChatGPT-DetectorModel=GPT4 Turbo2024.05 | 0.31 | |
| ChatGPT-DetectorModel=GPT42024.05 | 0.281 | |
| ChatGPT-DetectorSource Model=Qwen2 1.5B-it2024.05 | 0.239 | |
| ChatGPT-DetectorSource Model=OLMo 7B-sft2024.05 | 0.236 | |
| ChatGPT-DetectorModel=GPT3.5 Turbo2024.05 | 0.219 |