Understanding Latent Diffusability via Fisher Geometry

• URL: http://arxiv.org/abs/2604.02751v1
• Date: Fri, 03 Apr 2026 05:52:09 GMT
• Title: Understanding Latent Diffusability via Fisher Geometry
• Abstract summary: We quantify latent-space diffusability through the rate of change of the Minimum Mean Squared Error (MMSE)<n>Our framework decomposes this MMSE rate into contributions from Fisher Information (FI) and Fisher Information Rate (FIR)<n>Our analysis explicitly decouples latent geometric distortion into three measurable penalties: dimensional compression, tangential distortion, and curvature injection.
• Score: 27.84196444245208
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Diffusion models often degrade when trained in latent spaces (e.g., VAEs), yet the formal causes remain poorly understood. We quantify latent-space diffusability through the rate of change of the Minimum Mean Squared Error (MMSE) along the diffusion trajectory. Our framework decomposes this MMSE rate into contributions from Fisher Information (FI) and Fisher Information Rate (FIR). We demonstrate that while global isometry ensures FI alignment, FIR is governed by the encoder's local geometric properties. Our analysis explicitly decouples latent geometric distortion into three measurable penalties: dimensional compression, tangential distortion, and curvature injection. We derive theoretical conditions for FIR preservation across spaces, ensuring maintained diffusability. Experiments across diverse autoencoding architectures validate our framework and establish these efficient FI and FIR metrics as a robust diagnostic suite for identifying and mitigating latent diffusion failure.

Related papers

• Patch-PODiff-ViT: Structured Latent Diffusion with Patchwise POD for Super-Resolution and Uncertainty Quantification [2.075423143311858]
  We introduce Patch-PODiff-ViT, a structured latent diffusion framework in which the latent space is defined by patchwise Proper Orthogonal Decomposition (POD)<n>This yields low-dimensional, variance-ordered tokens that preserve spatial structure and enable efficient diffusion in a structured low-dimensional latent space with a Vision Transformer.<n>Across sea surface temperature, medical imaging, and natural images, the method achieves strong reconstruction with fewer parameters and lower memory, while producing well-calibrated spatial uncertainty that closely matches empirical ensembles.
  arXiv  Detail & Related papers  (2026-06-30T08:05:53Z)
• Disentangling Continuous-Time Latent Dynamics: Identifiability of Latent SDEs via Diffusion Shifts [60.95873409673775]
  identifiability in continuous-time latent differential equation models remains largely open.<n>We study additive-noise latent SDEs observed through an unknown nonlinear diffeomorphism.<n>We propose a two-stage estimator for latent disentanglement and optional graph recovery.
  arXiv  Detail & Related papers  (2026-06-26T16:18:28Z)
• Geometry Adaptive Counterfactual Distribution Learning with Diffusion-Guided Smoothing [4.727838288363022]
  Standard isotropic smoothing treats all ambient directions equally, leading to unfavorable scaling and unstable local inference.<n>We propose two diffusion-guided estimators based on semi-parametric debiasing: diffusion-informed smoothing for counterfactual densities and diffusion-informed score smoothing for counterfactual scores.
  arXiv  Detail & Related papers  (2026-05-25T13:02:56Z)
• High-Dimensional Latents Should Be Diagnosed Through Phase Structure [56.362776482614976]
  We study autoencoder and variational-autoencoder latent spaces through the lens of spin-glass theory.<n>We formalize a latent-space spin-glass dictionary for a fixed decoder.<n>We show that driving the latent system toward the edge-of-stability of the topological trivialization regime has concrete downstream consequences.
  arXiv  Detail & Related papers  (2026-05-23T06:17:23Z)
• Provably Learning Diffusion Models under the Manifold Hypothesis: Collapse and Refine [60.669081685261965]
  Diffusion models generate high-dimensional data with remarkable quality.<n>How their training efficiently learns the score function remains theoretically unexplained.<n>We instantiate this principle as Score-induced Latent Diffusion (SiLD)
  arXiv  Detail & Related papers  (2026-05-16T16:51:10Z)
• Oversmoothing as Representation Degeneracy in Neural Sheaf Diffusion [0.0]
  Neural Sheaf Diffusion (NSD) generalizes diffusion-based Graph Neural Networks by replacing scalar graph Laplacians with sheaf Laplacians.<n>We develop a quiver-theoretic interpretation of NSD by identifying cellular sheaves on graphs with representations of the associated incidence.<n>We show that direct-sum decompositions of the underlying incidence-quiver representation induce decompositions of the harmonic space reached in the diffusion limit.
  arXiv  Detail & Related papers  (2026-05-11T19:41:56Z)
• Information Fidelity in Tool-Using LLM Agents: A Martingale Analysis of the Model Context Protocol [69.11739400975445]
  We introduce the first theoretical framework for analyzing error accumulation in Model Context Protocol (MCP) agents.<n>We show that cumulative distortion exhibits linear growth and high-probability deviations bounded by $O(sqrtT)$.<n>Key findings include: semantic weighting reduces distortion by 80%, and periodic re-grounding approximately every 9 steps suffices for error control.
  arXiv  Detail & Related papers  (2026-02-10T21:08:53Z)
• Emergence of Distortions in High-Dimensional Guided Diffusion Models [11.774563966512707]
  We formalize the phenomenon of generative distortion defined as the mismatch between the CFG-induced sampling and the true conditional distribution.<n>We show that standard CFG schedules are incapable of preventing variance shrinkage.<n>We propose a theoretically motivated guidance schedule featuring a negative-guidance window, which mitigates loss of diversity while preserving class separability.
  arXiv  Detail & Related papers  (2026-01-31T13:19:45Z)
• Out-of-Distribution Detection in Molecular Complexes via Diffusion Models for Irregular Graphs [11.928558263824213]
  We present a probabilistic OOD detection framework for complex 3D graph data built on a diffusion model.<n>A single probability-flow ODE produces per-sample log-likelihoods, providing a principled typicality score for distribution shift.<n>We validate the approach on protein-ligand complexes and construct strict OOD datasets.
  arXiv  Detail & Related papers  (2025-12-20T17:56:15Z)
• Hessian Geometry of Latent Space in Generative Models [41.94295877935867]
  We present a novel method for analyzing the latent space geometry of generative models.<n>The method approximates the posterior distribution of latent variables given generated samples.<n>It is validated on the Ising and TASEP models, outperforming existing baselines in reconstructing thermodynamic quantities.
  arXiv  Detail & Related papers  (2025-06-12T12:17:40Z)
• Learning over von Mises-Fisher Distributions via a Wasserstein-like Geometry [0.0]
  We introduce a geometry-aware distance metric for the family of von Mises-Fisher (vMF) distributions.<n>Motivated by the theory of optimal transport, we propose a Wasserstein-like distance that decomposes the discrepancy between two vMF distributions into two interpretable components.
  arXiv  Detail & Related papers  (2025-04-19T03:38:15Z)
• Rectified Diffusion Guidance for Conditional Generation [94.83538269086613]
  We revisit the theory behind CFG and rigorously confirm that the improper combination coefficients (textiti.e.) brings about expectation shift the generative distribution.<n>We show that our approach enjoys a textbftextitform solution given the strength.<n> Empirical evidence on real-world data demonstrate the compatibility of our design with existing state-of-the-art diffusion models.
  arXiv  Detail & Related papers  (2024-10-24T13:41:32Z)
• Reconstructing Graph Diffusion History from a Single Snapshot [87.20550495678907]
  We propose a novel barycenter formulation for reconstructing Diffusion history from A single SnapsHot (DASH) We prove that estimation error of diffusion parameters is unavoidable due to NP-hardness of diffusion parameter estimation. We also develop an effective solver named DIffusion hiTting Times with Optimal proposal (DITTO)
  arXiv  Detail & Related papers  (2023-06-01T09:39:32Z)
• A Unifying and Canonical Description of Measure-Preserving Diffusions [60.59592461429012]
  A complete recipe of measure-preserving diffusions in Euclidean space was recently derived unifying several MCMC algorithms into a single framework. We develop a geometric theory that improves and generalises this construction to any manifold.
  arXiv  Detail & Related papers  (2021-05-06T17:36:55Z)
• On the Practicality of Differential Privacy in Federated Learning by Tuning Iteration Times [51.61278695776151]
  Federated Learning (FL) is well known for its privacy protection when training machine learning models among distributed clients collaboratively. Recent studies have pointed out that the naive FL is susceptible to gradient leakage attacks. Differential Privacy (DP) emerges as a promising countermeasure to defend against gradient leakage attacks.
  arXiv  Detail & Related papers  (2021-01-11T19:43:12Z)
• Manifold-Aware CycleGAN for High-Resolution Structural-to-DTI Synthesis [8.829738147738222]
  We propose a manifold-aware CycleGAN that learns the generation of high-resolution DTI from unpaired T1w images. Our method is able to generate realistic high-resolution DTI that can be used to compute diffusion-based metrics and potentially run fiber tractography algorithms.
  arXiv  Detail & Related papers  (2020-04-01T00:08:14Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.