In Situ Training of Implicit Neural Compressors for Scientific Simulations via Sketch-Based Regularization

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Simpson, Cooper, Becker, Stephen, Doostan, Alireza
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908974666743808
author Simpson, Cooper
Becker, Stephen
Doostan, Alireza
author_facet Simpson, Cooper
Becker, Stephen
Doostan, Alireza
contents Focusing on implicit neural representations, we present a novel in situ training protocol that employs limited memory buffers of full and sketched data samples, where the sketched data are leveraged to prevent catastrophic forgetting. The theoretical motivation for our use of sketching as a regularizer is presented via a simple Johnson-Lindenstrauss-informed result. While our methods may be of wider interest in the field of continual learning, we specifically target in situ neural compression using implicit neural representation-based hypernetworks. We evaluate our method on a variety of complex simulation data in two and three dimensions, over long time horizons, and across unstructured grids and non-Cartesian geometries. On these tasks, we show strong reconstruction performance at high compression rates. Most importantly, we demonstrate that sketching enables the presented in situ scheme to approximately match the performance of the equivalent offline method.
format Preprint
id arxiv_https___arxiv_org_abs_2511_02659
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle In Situ Training of Implicit Neural Compressors for Scientific Simulations via Sketch-Based Regularization
Simpson, Cooper
Becker, Stephen
Doostan, Alireza
Machine Learning
Artificial Intelligence
Computational Engineering, Finance, and Science
Numerical Analysis
Focusing on implicit neural representations, we present a novel in situ training protocol that employs limited memory buffers of full and sketched data samples, where the sketched data are leveraged to prevent catastrophic forgetting. The theoretical motivation for our use of sketching as a regularizer is presented via a simple Johnson-Lindenstrauss-informed result. While our methods may be of wider interest in the field of continual learning, we specifically target in situ neural compression using implicit neural representation-based hypernetworks. We evaluate our method on a variety of complex simulation data in two and three dimensions, over long time horizons, and across unstructured grids and non-Cartesian geometries. On these tasks, we show strong reconstruction performance at high compression rates. Most importantly, we demonstrate that sketching enables the presented in situ scheme to approximately match the performance of the equivalent offline method.
title In Situ Training of Implicit Neural Compressors for Scientific Simulations via Sketch-Based Regularization
topic Machine Learning
Artificial Intelligence
Computational Engineering, Finance, and Science
Numerical Analysis
url https://arxiv.org/abs/2511.02659