# Fixed-point acceleration on Neural Wasserstein manifolds

DevFeed: [Fixed-point acceleration on Neural Wasserstein manifolds](<https://devfeed.tech/articles/fixed-point-acceleration-on-neural-wasserstein-manifolds-123448.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.09723>)

Author: Vitalii Aksenov, Sebastian Gutierrez Hernandez

Published: 2026-11-09T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Neural Network](<https://devfeed.tech/topics/neural-network.md>)

Tags: [acceleration](<https://devfeed.tech/tags/acceleration.md>), [aggregation](<https://devfeed.tech/tags/aggregation.md>), [bayesian](<https://devfeed.tech/tags/bayesian.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [distribution](<https://devfeed.tech/tags/distribution.md>), [efficiency](<https://devfeed.tech/tags/efficiency.md>), [fixed](<https://devfeed.tech/tags/fixed.md>), [generative](<https://devfeed.tech/tags/generative.md>), [gradient](<https://devfeed.tech/tags/gradient.md>), [gradient-descent](<https://devfeed.tech/tags/gradient-descent.md>), [iteration](<https://devfeed.tech/tags/iteration.md>), [learning](<https://devfeed.tech/tags/learning.md>), [logistic-regression](<https://devfeed.tech/tags/logistic-regression.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [map](<https://devfeed.tech/tags/map.md>), [math-oc](<https://devfeed.tech/tags/math-oc.md>), [metric](<https://devfeed.tech/tags/metric.md>), [mixing](<https://devfeed.tech/tags/mixing.md>), [modeling](<https://devfeed.tech/tags/modeling.md>), [multimodal](<https://devfeed.tech/tags/multimodal.md>), [network](<https://devfeed.tech/tags/network.md>), [neural](<https://devfeed.tech/tags/neural.md>)

## AI overview

The paper accelerates natural-gradient optimization of neural pushforward probability measures with Riemannian Anderson mixing. The induced pullback Wasserstein metric makes natural-gradient descent a fixed-point iteration; mixing uses past residuals to capture approximate second-order information without forming Hessians. In this parameterized setting, Euclidean coordinates with a point-dependent metric simplify retraction and vector transport. Tests on multimodal sampling, PDE-based inverse problems, aggregation dynamics, Bayesian logistic regression, and generative modeling report fewer natural-gradient iterations and competitive performance against SGD and Adam in loss-gradient evaluations and wall-clock time. The evidence is limited to the abstract and does not specify task-level results.

## Source excerpt

arXiv:2611.09723v1 Announce Type: new Abstract: We propose a fixed-point acceleration method for optimization over probability measures, applicable to a variety of machine learning problems. We represent probability measures as pushforwards of a reference distribution through a neural network map and equip the parameter space with the induced pullback Wasserstein metric. This geometric structure allows natural gradient descent to be interpreted as a fixed-point iteration on a Riemannian manifold. We accelerate this iteration using Riemannian Anderson mixing method, which exploits historical residuals to capture approximate second-order information without forming Hessians. In the present parametric setting, the manifold is represented by Euclidean coordinates endowed with a point-dependent metric, simplifying retraction and vector transport. We evaluate the method on multimodal sampling, PDE-based inverse problems, aggregation dynamics, Bayesian logistic regression, and generative modeling. Across the tested tasks, Anderson mixing reduces the number of natural-gradient iterations and achieves competitive performance relative to SGD and Adam in terms of both loss-gradient evaluations and wall-clock time. These results suggest that fixed-point acceleration is an efficient tool for improving the computational efficiency of training on parametric Wasserstein methods.