English
Related papers

Related papers: Pulling back information geometry

200 papers

We present a novel geometric perspective on the latent space of diffusion models. We first show that the standard pullback approach, utilizing the deterministic probability flow ODE decoder, is fundamentally flawed. It provably forces…

Machine Learning · Computer Science 2026-02-27 Rafał Karczewski , Markus Heinonen , Alison Pouplin , Søren Hauberg , Vikas Garg

Latent space geometry provides a rigorous and empirically valuable framework for interacting with the latent variables of deep generative models. This approach reinterprets Euclidean latent spaces as Riemannian through a pull-back metric,…

Machine Learning · Statistics 2024-08-15 Stas Syrota , Pablo Moreno-Muñoz , Søren Hauberg

This paper presents a novel method for analyzing the latent space geometry of generative models, including statistical physics models and diffusion models, by reconstructing the Fisher information metric. The method approximates the…

Machine Learning · Computer Science 2025-06-13 Alexander Lobashev , Dmitry Guskov , Maria Larchenko , Mikhail Tamm

Deep generative models are universal tools for learning data distributions on high dimensional data spaces via a mapping to lower dimensional latent spaces. We provide a study of latent space geometries and extend and build upon previous…

Machine Learning · Computer Science 2019-02-07 Max F. Frenzel , Bogdan Teleaga , Asahi Ushio

Many deep generative models are defined as a push-forward of a Gaussian measure by a continuous generator, such as Generative Adversarial Networks (GANs) or Variational Auto-Encoders (VAEs). This work explores the latent space of such deep…

Machine Learning · Computer Science 2023-05-16 Thibaut Issenhuth , Ugo Tanielian , Jérémie Mary , David Picard

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…

Machine Learning · Statistics 2021-12-14 Georgios Arvanitidis , Lars Kai Hansen , Søren Hauberg

Latent variable models are powerful tools for learning low-dimensional manifolds from high-dimensional data. However, when dealing with constrained data such as unit-norm vectors or symmetric positive-definite matrices, existing approaches…

Machine Learning · Computer Science 2025-03-10 Leonel Rozo , Miguel González-Duque , Noémie Jaquier , Søren Hauberg

We investigate the geometrical structure of probabilistic generative dimensionality reduction models using the tools of Riemannian geometry. We explicitly define a distribution over the natural metric given by the models. We provide the…

Machine Learning · Statistics 2014-12-01 Alessandra Tosi , Søren Hauberg , Alfredo Vellido , Neil D. Lawrence

When dealing with a parametric statistical model, a Riemannian manifold can naturally appear by endowing the parameter space with the Fisher information metric. The geometry induced on the parameters by this metric is then referred to as…

Machine Learning · Statistics 2023-10-03 Florent Bouchard , Arnaud Breloy , Antoine Collas , Alexandre Renaux , Guillaume Ginolhac

Statistical inference more often than not involves models which are non-linear in the parameters thus leading to non-Gaussian posteriors. Many computational and analytical tools exist that can deal with non-Gaussian distributions, and…

General Relativity and Quantum Cosmology · Physics 2021-01-20 Eileen Giesel , Robert Reischke , Björn Malte Schäfer , Dominic Chia

Latent representations are an important theme in modern machine learning. Any network training with the notion of locality introduces a latent geometry which we can analyze with the help of differential geometry, specifically information…

High Energy Physics - Phenomenology · Physics 2026-05-07 Rebecca Maria Kuntz , Tilman Plehn , Björn Malte Schäfer , Benedikt Schosser , Sophia Vent

We propose a geometric latent-subspace framework for generative modeling of discrete data. Specifically, we introduce latent subspaces in the exponential parameter space of product manifolds of categorical distributions as a novel method…

Machine Learning · Statistics 2026-05-08 Daniel Gonzalez-Alvarado , Jonas Cassel , Stefania Petra , Christoph Schnörr

The Fisher-Rao metric from Information Geometry is related to phase transition phenomena in classical statistical mechanics. Several studies propose to extend the use of Information Geometry to study more general phase transitions in…

Statistical Mechanics · Physics 2016-05-04 Omri Har Shemesh , Rick Quax , Alfons G. Hoekstra , Peter M. A. Sloot

Continuous diffusion and flow models are attractive for non-autoregressive text generation because they can update all positions in parallel. A major difficulty is the interface between continuous latent states and discrete tokens. This…

Computation and Language · Computer Science 2026-05-18 De Shuai Zhang

Generative adversarial networks (GANs) have emerged as a powerful unsupervised method to model the statistical patterns of real-world data sets, such as natural images. These networks are trained to map random inputs in their latent space…

Machine Learning · Computer Science 2021-03-19 Binxu Wang , Carlos R. Ponce

Deep generative models like variational autoencoders approximate the intrinsic geometry of high dimensional data manifolds by learning low-dimensional latent-space variables and an embedding function. The geometric properties of these…

Computer Vision and Pattern Recognition · Computer Science 2019-02-20 Ankita Shukla , Shagun Uppal , Sarthak Bhagat , Saket Anand , Pavan Turaga

In this paper, we advocate the adoption of metric preservation as a powerful prior for learning latent representations of deformable 3D shapes. Key to our construction is the introduction of a geometric distortion criterion, defined…

Machine Learning · Computer Science 2020-12-14 Luca Cosmo , Antonio Norelli , Oshri Halimi , Ron Kimmel , Emanuele Rodolà

Given data, deep generative models, such as variational autoencoders (VAE) and generative adversarial networks (GAN), train a lower dimensional latent representation of the data space. The linear Euclidean geometry of data space pulls back…

Computer Vision and Pattern Recognition · Computer Science 2018-05-22 Line Kuhnel , Tom Fletcher , Sarang Joshi , Stefan Sommer

Variational Autoencoder is typically understood from the perspective of probabilistic inference. In this work, we propose a new geometric reinterpretation which complements the probabilistic view and enhances its intuitiveness. We…

Machine Learning · Computer Science 2025-08-22 Songxuan Shi

Quantum state tomography faces exponential scaling with system size, while recent neural network approaches achieve polynomial scaling at the cost of losing the geometric structure of quantum state space. We introduce geometric latent space…

Quantum Physics · Physics 2025-12-19 S. M. Yousuf Iqbal Tomal , Abdullah Al Shafin
‹ Prev 1 2 3 10 Next ›