List your company
ResearchFormerly WizWand

Classification on HIGGS (test)

79.1AUCReLU
65.37268.93672.576.064Jan 10, 2026Feb 3, 2026Feb 28, 2026Mar 25, 2026Apr 19, 2026May 14, 2026Jun 8, 2026
Updated 4mo ago

Evaluation results

MethodLinks
ReLUArchitecture=Heavy, Params=26,6012026.0179.171.80.179
ReLUArchitecture=Light, Params=4,1612026.0178.4710.217
GELUArchitecture=Light, Params=4,1612026.0178.170.90.216
Hybrid (Specialist)Architecture=Light, Params=4,1612026.0177.770.70.215
SiLUArchitecture=Light, Params=4,1612026.017770.10.213
Gaussian RFFM=8192, D=8192, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0676——
Gaussian RFFM=4096, D=4096, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0675.5——
Gaussian RFFM=2048, D=2048, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0674.9——
RAY (sketch)M=8192, D=52, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0674.9——
RAY (sketch)M=4096, D=26, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0674.3——
exact modulationM=4096, D=9, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0674.1——
Gaussian RFFM=1024, D=1024, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0673.7——
exact modulationM=8192, D=18, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0673.6——
RAY (sketch)M=2048, D=13, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0673.4——
exact modulationM=2048, D=4, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0673.2——
Gaussian RFFM=512, D=512, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0671.9——
RAY (sketch)M=1024, D=6, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0671.9——
RAY (sketch)M=512, D=3, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0669.9——
exact modulationM=1024, D=2, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0669.8——
Linear baseline2026.0668.2——
exact modulationM=512, D=1, optimizer=Streaming Adam, batch size=8192, learning rate=0.022026.0665.9——