AI-Generated Text Detection on RAID (test)

78.4TP @ 20% Error ThresholdBinoculars

Evaluation results

Method
Binoculars2025.05	78.4	79.03	76.47	82.02	74.53	83.03	70.17	81.5	67.5	80.12	64.17	77.94
Binoculars2025.05	78.4	79.03	76.47	82.02	74.53	83.03	70.17	81.5	67.5	80.12	64.17	77.94
Binoculars2025.05	78.4	79.03	76.47	82.02	74.53	83.03	70.17	81.5	67.5	80.12	64.17	77.94
Binoculars2025.05	78.17	79.17	76.57	82.21	74.83	83.19	72.37	82.78	70.33	81.93	66.07	79.25
Binoculars2025.05	78.17	79.17	76.57	82.21	74.83	83.19	72.37	82.78	70.33	81.93	66.07	79.25
Binoculars2025.05	78.17	79.17	76.57	82.21	74.83	83.19	72.37	82.78	70.33	81.93	66.07	79.25
Fast-DetectGPT2025.05	78.13	78.88	76.03	81.77	73.43	82.4	68.27	80.2	64.73	78.09	63.6	77.56
Fast-DetectGPT2025.05	78.13	78.88	76.03	81.77	73.43	82.4	68.27	80.2	64.73	78.09	63.6	77.56
Fast-DetectGPT2025.05	78.13	78.88	76.03	81.77	73.43	82.4	68.27	80.2	64.73	78.09	63.6	77.56
Fast-DetectGPT2025.05	77.97	78.74	75.23	81.22	71.67	81.13	66.7	79.08	64.7	78.09	63.1	77.14
Fast-DetectGPT2025.05	77.97	78.74	75.23	81.22	71.67	81.13	66.7	79.08	64.7	78.09	63.1	77.14
Fast-DetectGPT2025.05	77.97	78.74	75.23	81.22	71.67	81.13	66.7	79.08	64.7	78.09	63.1	77.14
OpenAI-D2025.05	74.47	76.57	63.8	73.4	56.07	69.62	45.63	61.82	39.93	56.67	17	28.94
OpenAI-D2025.05	74.47	76.57	63.8	73.4	56.07	69.62	45.63	61.82	39.93	56.67	17	28.94
OpenAI-D2025.05	74.47	76.57	63.8	73.4	56.07	69.62	45.63	61.82	39.93	56.67	17	28.94
OpenAI-D2025.05	73.83	76.33	62.63	72.87	54.77	68.83	47.07	63.13	41.93	58.66	33.47	49.93
OpenAI-D2025.05	73.83	76.33	62.63	72.87	54.77	68.83	47.07	63.13	41.93	58.66	33.47	49.93
OpenAI-D2025.05	73.83	76.33	62.63	72.87	54.77	68.83	47.07	63.13	41.93	58.66	33.47	49.93
Log-Rank2025.05	66.1	71.01	54.1	65.91	40.63	55.8	28.83	44.08	22.67	36.66	19.77	32.88
Log-Rank2025.05	66.1	71.01	54.1	65.91	40.63	55.8	28.83	44.08	22.67	36.66	19.77	32.88
Log-Rank2025.05	66.1	71.01	54.1	65.91	40.63	55.8	28.83	44.08	22.67	36.66	19.77	32.88
Log-Rank2025.05	64.97	70.26	54.73	66.6	46.47	61.61	39.43	55.7	34.23	50.59	29.4	45.23
Log-Rank2025.05	64.97	70.26	54.73	66.6	46.47	61.61	39.43	55.7	34.23	50.59	29.4	45.23
Log-Rank2025.05	64.97	70.26	54.73	66.6	46.47	61.61	39.43	55.7	34.23	50.59	29.4	45.23
Likelihood2025.05	64.37	69.82	52.47	64.59	38.7	53.86	24.8	39.12	19.37	32.18	14	24.46
Likelihood2025.05	64.37	69.82	52.47	64.59	38.7	53.86	24.8	39.12	19.37	32.18	14	24.46
Likelihood2025.05	64.37	69.82	52.47	64.59	38.7	53.86	24.8	39.12	19.37	32.18	14	24.46
Likelihood2025.05	62.93	68.87	53.57	65.59	44.9	60	36.5	52.63	32.17	48.2	26.63	41.83
Likelihood2025.05	62.93	68.87	53.57	65.59	44.9	60	36.5	52.63	32.17	48.2	26.63	41.83
Likelihood2025.05	62.93	68.87	53.57	65.59	44.9	60	36.5	52.63	32.17	48.2	26.63	41.83
DetectGPT2025.05	60.1	67.06	48.37	61.51	35.2	50.64	21.2	34.54	12.67	22.38	7.47	13.86
DetectGPT2025.05	60.1	67.06	48.37	61.51	35.2	50.64	21.2	34.54	12.67	22.38	7.47	13.86
DetectGPT2025.05	60.1	67.06	48.37	61.51	35.2	50.64	21.2	34.54	12.67	22.38	7.47	13.86
DetectGPT2025.05	58.57	65.62	46.53	59.45	34.5	49.46	21.2	34.41	14.67	25.36	8.8	16.1
DetectGPT2025.05	58.57	65.62	46.53	59.45	34.5	49.46	21.2	34.41	14.67	25.36	8.8	16.1
DetectGPT2025.05	58.57	65.62	46.53	59.45	34.5	49.46	21.2	34.41	14.67	25.36	8.8	16.1
Entropy2025.05	42.33	52.3	29.33	42.37	23.53	36.61	17.43	29.11	14.73	25.43	12.27	21.74
Entropy2025.05	42.33	52.3	29.33	42.37	23.53	36.61	17.43	29.11	14.73	25.43	12.27	21.74
Entropy2025.05	42.33	52.3	29.33	42.37	23.53	36.61	17.43	29.11	14.73	25.43	12.27	21.74
Entropy2025.05	40.03	50.06	22.37	33.8	17.9	29.14	15.57	26.48	14.13	24.54	11.77	20.96
Entropy2025.05	40.03	50.06	22.37	33.8	17.9	29.14	15.57	26.48	14.13	24.54	11.77	20.96
Entropy2025.05	40.03	50.06	22.37	33.8	17.9	29.14	15.57	26.48	14.13	24.54	11.77	20.96