Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

DetectRL

Benchmarks

Task NameDataset NameSOTA ResultTrend
Machine-generated text detectionDetectRL Training Text: Llama-2-70b (test)
Detection Score (Llama-2-70b)90.2
12
Machine-Generated Text DetectionDetectRL (test)
Detection Score (Llama-2-70b)50.56
12
Machine-Generated Text DetectionDetectRL Google-PaLM (train)
TPR@FPR-1% (Llama-2-70b)50.58
12
Machine-Generated Text DetectionDetectRL Training Text: ChatGPT
TPR@FPR-1% (Llama-2-70b)50.66
12
Machine-generated text detectionDetectRL Google-PaLM
AUROC77.8
6
Machine-generated text detectionDetectRL Llama-2-70b
AUROC0.8122
6
Showing 6 of 6 rows