H by rule (mean over the two training seeds; synthetic = mean of the 5 LODO datasets); delta vs HMM 0.998 of the same replicate [all4] fam AASD Carta2026 synthetic dAASD dCarta2026 dsynthetic rule HMM 0.95 0.6486 0.6482 0.7903 0.0106 0.0094 -0.0078 HMM 0.998 0.6380 0.6388 0.7981 0.0000 0.0000 0.0000 V1 0.6374 0.5703 0.5165 -0.0006 -0.0685 -0.2816 V2 0.6371 0.5860 0.8109 -0.0009 -0.0527 0.0127 [s01] fam AASD Carta2026 synthetic dAASD dCarta2026 dsynthetic rule HMM 0.95 0.6476 0.6506 0.7863 0.0122 0.0059 -0.0080 HMM 0.998 0.6354 0.6448 0.7943 0.0000 0.0000 0.0000 V1 0.6361 0.5702 0.5160 0.0007 -0.0746 -0.2783 V2 0.6356 0.5829 0.8062 0.0002 -0.0618 0.0119 [s23] fam AASD Carta2026 synthetic dAASD dCarta2026 dsynthetic rule HMM 0.95 0.6513 0.6392 0.7886 0.0117 0.0070 -0.0076 HMM 0.998 0.6396 0.6322 0.7962 0.0000 0.0000 0.0000 V1 0.6389 0.5698 0.5008 -0.0007 -0.0624 -0.2954 V2 0.6374 0.5843 0.8092 -0.0022 -0.0480 0.0130 Per training seed (all4): H of AASD / Carta / synthetic fam AASD Carta2026 synthetic rule seed HMM 0.95 -1 0.6486 0.6482 0.7903 HMM 0.998 -1 0.6380 0.6388 0.7981 V1 0 0.6381 0.5743 0.5160 1 0.6366 0.5664 0.5171 V2 0 0.6354 0.5840 0.8124 1 0.6388 0.5881 0.8094 Per dataset (all4), V1 vs HMM 0.998 (H / Acc / SDT) H Acc SDT rule HMM 0.998 V1 HMM 0.998 V1 HMM 0.998 V1 ds AASD 0.638 0.637 52.805 52.737 13.438 13.216 AVGC 0.726 0.386 68.854 50.697 15.430 69.553 Carta2026 0.639 0.570 59.241 57.413 19.375 34.469 DTU 0.842 0.810 75.833 69.126 3.238 0.989 KUL 0.809 0.404 74.832 53.232 6.664 67.972 UHD 0.791 0.382 73.547 52.474 7.484 71.491 USTC 0.823 0.601 73.455 58.846 2.723 32.668 Per dataset (all4), V2 vs HMM 0.998 (H / Acc / SDT) H Acc SDT rule HMM 0.998 V2 HMM 0.998 V2 HMM 0.998 V2 ds AASD 0.638 0.637 52.805 52.702 13.438 13.326 AVGC 0.726 0.749 68.854 70.423 15.430 12.539 Carta2026 0.639 0.586 59.241 53.884 19.375 26.156 DTU 0.842 0.841 75.833 74.801 3.238 2.293 KUL 0.809 0.829 74.832 76.569 6.664 5.176 UHD 0.791 0.808 73.547 75.837 7.484 7.031 USTC 0.823 0.828 73.455 74.826 2.723 4.062