Backdoor detection on CIFAR-10 imbalanced µ=0.9, ρ=2 (test) benchmark leaderboard · SOTA2 Research

Backdoor detection on CIFAR-10 imbalanced µ=0.9, ρ=2 (test)

96.7Badnets TPRRPP

Evaluation Results

Method	Links
RPP2026.01		96.7	3.2	89.5	9.3	89.9	12.9	93	4.7	89	10.9	90.6	8.8	93.8	24.7	87.4	4.9	86.7	11.2	95.6	16.9
TED2026.01		96.2	4.8	90.9	8.2	76.5	12.1	88	12.9	90.3	18.4	88.4	11.9	89.9	13.3	86	13.6	83.9	14	90.3	14
SCALE-UP2026.01		95.8	44	75.7	28.4	55.7	38.9	99	40.1	90.6	29.8	85.4	21.1	32.6	4.6	76.7	69.2	72.4	39.7	60.5	17.2
BBCaL2026.01		95.7	18.4	94.2	19.3	72.8	13	89.4	19.6	93.1	17.7	81.8	20.3	80.4	17.1	70.1	24.4	67.3	20.5	84	16.1
IBD-PSC2026.01		95.3	6.8	93.6	9.5	90.3	10	85.2	12.6	90.3	13.6	88.5	12.1	64.3	14.4	85.3	9.7	85.6	14.9	90.5	22.9
STRIP2026.01		90.2	18.9	77.8	26.3	20.7	10.5	88.6	30.7	40.2	22.8	10.5	33.2	17.6	10.3	85.4	24	80.9	26.3	0	0.8
MSPC2026.01		88.2	45.1	90.7	20.4	59.8	25.9	80.1	38.9	78.3	18.8	89.8	9.6	67.6	33.8	64.3	34.8	70.3	27.8	77.5	27.1
CT2026.01		75.1	12.6	84.3	10.2	82.1	3.2	90.1	37.6	91.6	18.3	92.5	15.7	75.3	10.4	68.7	0.6	77.2	8.5	10.6	16.3
RE2026.01		73.9	17	72.7	20.5	56.5	16.9	38.7	11	41.5	16.4	50.4	12.3	52.9	11.7	57.8	17.5	50.4	20.8	10.1	12.3
AC2026.01		56.1	16.1	42.7	13.5	24	49.6	66.7	16.3	44.1	9.6	58.3	6.7	32.1	20.2	20.7	17.3	21.3	12.6	3.6	21.6
SS2026.01		55.6	1.2	57.9	9.6	54.7	2	62.1	25.2	39.5	1.9	60	8.2	60	2.3	11.7	2.8	16.1	3.9	8.3	7.7
ASSET2026.01		53.3	31.4	55.1	27.6	71.9	37.1	62.2	36.4	80.4	15.6	81.1	16.4	30.5	29.4	32.9	36.9	65.9	22.8	89	6.2
ABL2026.01		23.6	2.6	36.2	8	15.2	0.3	10.9	3.2	33.7	0.6	60.3	11.5	4.9	1.3	12.1	0.6	10.3	4.7	0	0.5