Malware Detection on BODMAS, PMML, and DikeDataset (test) benchmark leaderboard · SOTA2 Research

Malware Detection on BODMAS, PMML, and DikeDataset (test)

0.9623Benign PrecisionMoE-Top-2 Experts

Evaluation Results

Method
MoE-Top-2 Experts2026.02	0.9623	0.8644	0.9107	0.9223	0.9794	0.95	0.9359
MoE-Uniform Aggregation2026.02	0.9273	0.8644	0.8947	0.9208	0.9588	0.9394	0.9231
MoE-Top-2 Experts w/o LB2026.02	0.9259	0.8475	0.885	0.9118	0.9588	0.9347	0.9167
MoE-Top-1 Expert2026.02	0.8689	0.8983	0.8833	0.9368	0.9175	0.9271	0.9103
Consistency framework (GIN)2026.02	0.8664	0.7797	0.807	0.8713	0.9072	0.8889	0.859
MalGNE framework (GAT)2026.02	0.8621	0.8475	0.8547	0.9082	0.9175	0.9128	0.891
Consistency framework (GraphSAGE)2026.02	0.85	0.8644	0.8571	0.9167	0.9072	0.9119	0.891
MoE-Temperature Softmax2026.02	0.8387	0.8814	0.8595	0.9255	0.8969	0.911	0.891
MalGNE framework (GCN)2026.02	0.8197	0.8475	0.8333	0.9053	0.8866	0.8958	0.8718
Consistency framework (GAT)2026.02	0.8197	0.8475	0.8333	0.9053	0.8866	0.8958	0.8718
Consistency framework (GCN)2026.02	0.8167	0.8305	0.8235	0.8958	0.8866	0.8912	0.8654