The Surprising Effectiveness of Equivariant Models in Domains with Latent Symmetry. (arXiv:2211.09231v1 [cs.LG])

Extensive work has demonstrated that equivariant neural networks can
significantly improve sample efficiency and generalization by enforcing an
inductive bias in the network architecture. These applications typically assume
that the domain symmetry is fully described by explicit transformations of the
model inputs and outputs. However, many real-life applications contain only
latent or partial symmetries which cannot be easily described by simple
transformations of the input. In these cases, it is necessary to learn symmetry
in the environment instead of imposing …

arxiv symmetry