List your company

Modular arithmetic prediction on MNIST modular arithmetic Unseen operations (test)

99.46Zero-Shot Op AccuracyResNet w/ multi-frequency rotation (MFR)
-1.461624.739250.9477.1408May 31, 2026

Evaluation results

MethodLinks
ResNet w/ multi-frequency rotation (MFR)Backbone=ResNet, Training Objective=BRo-JEPA, Rotation type (SFR/MFR)=MFR, Embedding type=block rotation2026.0599.4699.4670.9
ResNet18 w/ compositional consistency lossBackbone=ResNet18, Training Objective=JEPA, Compositional consistency loss=true, Embedding type=additive2026.0599.3199.3178.39
ResNet w/ single-frequency rotation (SFR)Backbone=ResNet, Training Objective=BRo-JEPA, Rotation type (SFR/MFR)=SFR, Embedding type=block rotation2026.0598.9398.9369.22
MLP w/ compositional consistency lossBackbone=MLP, Training Objective=JEPA, Compositional consistency loss=true, Embedding type=additive2026.0597.7797.8697.81
MLP w/ multi-frequency rotation (MFR)Backbone=MLP, Training Objective=BRo-JEPA, Rotation type (SFR/MFR)=MFR, Embedding type=block rotation2026.0597.2497.2197.23
MLP w/ single-frequency rotation (SFR)Backbone=MLP, Training Objective=BRo-JEPA, Rotation type (SFR/MFR)=SFR, Embedding type=block rotation2026.0594.2694.2394.28
MLP w/ multi-frequency rotation (MFR)Backbone=MLP, Training Objective=Supervised, Rotation type (SFR/MFR)=MFR, Embedding type=block rotation2026.0556.5—53.59
ResNet18 w/ multi-frequency rotation (MFR)Backbone=ResNet18, Training Objective=Supervised, Rotation type (SFR/MFR)=MFR, Embedding type=block rotation2026.0554.92—50.41
MLP w/ single-frequency rotation (SFR)Backbone=MLP, Training Objective=Supervised, Rotation type (SFR/MFR)=SFR, Embedding type=block rotation2026.0513.92—14.22
ResNet18 w/ single-frequency rotation (SFR)Backbone=ResNet18, Training Objective=Supervised, Rotation type (SFR/MFR)=SFR, Embedding type=block rotation2026.0510.1—12.78
MLPBackbone=MLP, Training Objective=JEPA, Embedding type=additive2026.059.9597.9310.44
ResNet18Backbone=ResNet18, Training Objective=JEPA, Embedding type=additive2026.059.7499.479.03
MLPBackbone=MLP, Training Objective=Supervised, Embedding type=additive2026.055.93—1.8
ResNet18Backbone=ResNet18, Training Objective=Supervised, Embedding type=additive2026.052.42—0.11