Saved in:
Bibliographic Details
Main Authors: Thaler, Stephan, Fuchs, Paul, Cukarska, Ana, Zavadlav, Julija
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2505.11190
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915289922273280
author Thaler, Stephan
Fuchs, Paul
Cukarska, Ana
Zavadlav, Julija
author_facet Thaler, Stephan
Fuchs, Paul
Cukarska, Ana
Zavadlav, Julija
contents We present JaxSGMC, an application-agnostic library for stochastic gradient Markov chain Monte Carlo (SG-MCMC) in JAX. SG-MCMC schemes are uncertainty quantification (UQ) methods that scale to large datasets and high-dimensional models, enabling trustworthy neural network predictions via Bayesian deep learning. JaxSGMC implements several state-of-the-art SG-MCMC samplers to promote UQ in deep learning by reducing the barriers of entry for switching from stochastic optimization to SG-MCMC sampling. Additionally, JaxSGMC allows users to build custom samplers from standard SG-MCMC building blocks. Due to this modular structure, we anticipate that JaxSGMC will accelerate research into novel SG-MCMC schemes and facilitate their application across a broad range of domains.
format Preprint
id arxiv_https___arxiv_org_abs_2505_11190
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle JaxSGMC: Modular stochastic gradient MCMC in JAX
Thaler, Stephan
Fuchs, Paul
Cukarska, Ana
Zavadlav, Julija
Computation
Applications
Machine Learning
We present JaxSGMC, an application-agnostic library for stochastic gradient Markov chain Monte Carlo (SG-MCMC) in JAX. SG-MCMC schemes are uncertainty quantification (UQ) methods that scale to large datasets and high-dimensional models, enabling trustworthy neural network predictions via Bayesian deep learning. JaxSGMC implements several state-of-the-art SG-MCMC samplers to promote UQ in deep learning by reducing the barriers of entry for switching from stochastic optimization to SG-MCMC sampling. Additionally, JaxSGMC allows users to build custom samplers from standard SG-MCMC building blocks. Due to this modular structure, we anticipate that JaxSGMC will accelerate research into novel SG-MCMC schemes and facilitate their application across a broad range of domains.
title JaxSGMC: Modular stochastic gradient MCMC in JAX
topic Computation
Applications
Machine Learning
url https://arxiv.org/abs/2505.11190