REVIEW 9 cited by
Latent Space Oddity: on the Curvature of Deep Generative Models
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Signed reviews
read the original abstract
Deep generative models provide a systematic way to learn nonlinear data distributions, through a set of latent variables and a nonlinear "generator" function that maps latent points into the input space. The nonlinearity of the generator imply that the latent space gives a distorted view of the input space. Under mild conditions, we show that this distortion can be characterized by a stochastic Riemannian metric, and demonstrate that distances and interpolants are significantly improved under this metric. This in turn improves probability distributions, sampling algorithms and clustering in the latent space. Our geometric analysis further reveals that current generators provide poor variance estimates and we propose a new generator architecture with vastly improved variance estimates. Results are demonstrated on convolutional and fully connected variational autoencoders, but the formalism easily generalize to other deep generative models.
Forward citations
Cited by 9 Pith papers
-
Learning Transfers: Kan Extensions for Neural Invariants
Transfer quality is measured against the left Kan extension of the source invariant along the task-change functor; for one-parameter persistence this equals the bottleneck distance exactly.
-
Stable Autoregressive Speech Generation with Low-Frame-Rate High-Dimensional Continuous Tokens
Autoregressive TTS from 8-Hz, 768-dimensional continuous tokens works when the tokenizer shapes its latent space with a low-dimensional core and an energy hierarchy, and the generator separates guidance into local, se...
-
Riemannian Geometry for Pre-trained Language Model Embeddings
Aggregating per-token pullback metrics via the Fréchet mean on the SPD manifold outperforms Euclidean mean pooling for sentence classification, with most of the gain attributable to geometric aggregation rather than l...
-
Privacy-Utility Trade-off in Data Publication: A Bilevel Optimization Framework with Curvature-Guided Perturbation
A bilevel optimization framework with curvature-guided geodesic perturbation reduces membership inference attack success while preserving downstream classification accuracy and sample quality.
-
Hessian Geometry of Latent Space in Generative Models
A method to recover the Hessian/Fisher metric of generative model latent spaces is validated on Ising and TASEP and applied to claim fractal phase transitions in diffusion models.
-
Your Data Manifold is Secretly a Reward Model: Shell-LCC for Text-to-Video Generation
Shell-LCC models the high-quality data manifold as an isotropic shell to derive cost-free reward signals that improve realism and high-frequency details in text-to-video generation.
-
Geometric flow regularization in latent spaces for smooth dynamics with the efficient variations of curvature
Curvature-flow-regularized latent spaces improve mean out-of-distribution errors for Burger's equation relative to a plain autoencoder, but the flows are heuristic and the evidence is limited.
-
Ruled by the Representation Space: On the University's Embrace of Large Language Models
Universities that adopt LLMs without critical scrutiny let hidden, corporate-shaped norms encoded in the models' representation space govern teaching and research.
-
Machine Learning is Good for Physics - and Vice Versa
A perspective essay arguing that AI should be integrated into fundamental physics while preserving the field's statistical and theory-based standards, and that physics can enrich machine learning.
Discussion (0). Continue with ORCID to comment.