Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

LGT Detection on WritingPrompts small Fast-DetectGPT benchmark (test)

99.9AUROC

Fast-DetectGPT

19.40440.30261.282.098May 27, 2024
Updated 1mo ago

Evaluation Results

MethodLinks
2024.05
99.9
2024.05
99.8
2024.05
99.8
2024.05
99.5
2024.05
99.5
2024.05
99.4
2024.05
99.2
2024.05
98.8
2024.05
97.6
2024.05
97.4
2024.05
97.1
2024.05
96.1
2024.05
95.8
2024.05
95.5
2024.05
94.7
2024.05
94.2
2024.05
93.8
2024.05
93.5
2024.05
92.9
2024.05
92.8
2024.05
92.5
2024.05
91.9
2024.05
91.8
2024.05
90.1
2024.05
89.6
2024.05
86.5
2024.05
86.4
2024.05
85.5
2024.05
81.1
2024.05
80.9
2024.05
80.8
2024.05
80.7
2024.05
78
2024.05
78
2024.05
77.9
2024.05
77.8
2024.05
77.5
2024.05
75.8
2024.05
75
2024.05
72.7
2024.05
72.3
2024.05
71.5
2024.05
70.9
2024.05
70.3
2024.05
70.2
2024.05
65.7
2024.05
63
2024.05
50.7
2024.05
48
2024.05
45.3
2024.05
28.1
2024.05
27.5
2024.05
24.1
2024.05
22.5