Representation Learning on MNIST non-uniform benchmark leaderboard · SOTA2 Research

Representation Learning on MNIST non-uniform

91.4k-NN AccuracyGRAE

Evaluation Results

Method	Links
GRAE2026.04		91.4	0.106	0.0484	0.0052	0.0031
DN2026.04		88.1	0.0719	0.041	0.0042	0.0055
TAE2026.04		87.7	0.0335	0.0223	0.0013	0.0056
BLAE2026.04		86.5	0.0289	0.0166	0.0012	0.0039
SPAE2026.04		82.2	0.0559	0.0611	0.0054	0.0078
CAE2026.04		75.8	0.148	0.112	0.0046	0.0085
Vanilla AE2026.04		74.9	0.276	0.148	0.0077	0.0097
IRAE2026.04		73.6	0.0725	0.0793	0.0052	0.0088
GGAE2026.04		71.5	0.263	0.152	0.0089	0.0098
GAE2026.04		68.2	0.148	0.0922	0.0041	0.0145