Push-forwards of Gaussian or log-concave latent variables through Lipschitz neural networks are always sub-Gaussian or sub-exponential, so common deep generative models cannot generate heavy-tailed distributions.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
On the Statistical Capacity of Deep Generative Models
Push-forwards of Gaussian or log-concave latent variables through Lipschitz neural networks are always sub-Gaussian or sub-exponential, so common deep generative models cannot generate heavy-tailed distributions.