Flopping-equivariant versions of ResMLP, ViT and ConvNeXt use half the FLOPs of the originals and match their ImageNet-1K accuracy at large model sizes.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Flopping for FLOPs: Leveraging equivariance for computational efficiency
Flopping-equivariant versions of ResMLP, ViT and ConvNeXt use half the FLOPs of the originals and match their ImageNet-1K accuracy at large model sizes.