A two-layer MNIST study argues that networks prefer distance-based representations, but the evidence is mostly architecture-specific and the OffsetL2 layer resembles known RBF units.
The mnist database of handwritten digit images for machine learning research [best of the web]
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Neural Networks Learn Distance Metrics
A two-layer MNIST study argues that networks prefer distance-based representations, but the evidence is mostly architecture-specific and the OffsetL2 layer resembles known RBF units.