MATH · IN · MODELS

Moschella et al. (2022) image (variational) autoencoder

Sapienza University of Rome

Structures found in this family (1)

Platonic Representation Hypothesis

By model (1)

Moschella et al. (2022) convolutional (V)AE for MNIST/F-MNIST/CIFAR-10/CIFAR-100
Platonic Representation Hypothesis

Observations (1)

Re-expressing embeddings as cosine similarity to a fixed anchor set yields a rotation/reflection/rescaling-invariant coordinate system in which decoders and classifiers trained on one independently-trained encoder can be stitched, with zero fine-tuning, onto a completely different independently-trained encoder's representations of the same domain

Papers

Relative Representations Enable Zero-Shot Latent Space Communication (2022)