AI-Generated Text Detection on mixed-source AI -> Human GPT-2, GPT-Neo, GPT-J, LLaMa, GPT-3 benchmark leaderboard · SOTA2 Research

AI-Generated Text Detection on mixed-source AI -> Human GPT-2, GPT-Neo, GPT-J, LLaMa, GPT-3

96.5Overall AccuracySeqXGPT

Evaluation Results

Method	Links
SeqXGPT2024.06		96.5	65	32.64	1,867.81
RDrop2024.06		96.5	69	28.5	1,819.95
SeqXGPT2024.06		96.25	67.5	29.87	1,880.9
Flooding2024.06		96	60.5	36.98	1,800.01
DeBERTa2024.06		95.75	67.25	29.77	1,852.7
Flooding2024.06		95.75	56.75	40.73	1,705.23
RDrop2024.06		95.75	69.5	27.42	1,817.78
RoBERTa2024.06		95.5	64.5	32.46	1,840.19
DeBERTa2024.06		95.5	54.5	42.93	1,764.94
SCRN2024.06		95.5	89.25	6.54	2,043.44
SCRN2024.06		95	87	8.42	1,986.98
RoBERTa2024.06		94.25	63.5	32.63	1,784.96
RanMASK2024.06		94	60	36.17	1,784.11
RMLM2024.06		91.25	73.5	19.45	1,933.23
RMLM2024.06		91	69	24.18	1,879.96
BERT2024.06		90.5	1	98.9	1,204.52
BERT2024.06		90.25	30	66.76	1,509.98
RanMASK2024.06		90	65.5	27.22	1,749.92
GLTR2024.06		76.5	0	100	1,260.99
Log-Rank2024.06		73.5	0.5	99.32	1,286.24
Log-Likelihood2024.06		72	0.5	99.31	1,281.86
GLTR2024.06		72	9.5	86.81	1,273.32
Log-Rank2024.06		68	28.25	58.46	1,491.72
Log-Likelihood2024.06		67	27	59.7	1,474.88
Entropy2024.06		63	0	100	1,239.29
Entropy2024.06		59.25	13.75	76.79	1,317.84