Table 5.

This table shows the arithmetic mean of the average accuracies across all 10 tasks (in percent) for each model (columns) in each experiment (rows).

Dataset	GDM 6,000	GDM 12,000	GDM 28,000	A-SOINN+	(ncc)A-SOINN+
B1	73.95 ± 1.78	83.88 ± 0.66	92.60 ± 1.31	93.07 ± 3.47	95.71 ± 1.83
B2	73.65 ± 1.68	81.95 ± 2.02	89.86 ± 2.21	90.7 ± 3.62	90.06 ± 2.94
B3	72.09 ± 1.58	80.56 ± 2.08	84.88 ± 1.73	88.51 ± 2.21	88.28 ± 2.00
B4	70.96 ± 2.86	78.25 ± 2.63	80.69 ± 2.74	81.07 ± 6.30	86.93 ± 1.54
CORe50	85.58 ± 3.02	98.73 ± 0.97	99.19 ± 0.46	97.26 ± 1.76	99.33 ± 0.30

The gray shaded entries represent the highest values for each experiment.