# Quantifying the Impact of Lossy Compression on Neural Generative Surrogate Modeling

arXiv: https://arxiv.org/abs/2606.15959 · HTML: https://mlsanity.com/paper/2606.15959/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 39 references that are in the map.
- First posted on arXiv 14 Jun 2026.

## Builds on (most cited first, 21 of 21)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification](https://mlsanity.com/paper/1502.01852/) (2015)
- [Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks](https://mlsanity.com/paper/1511.06434/) (2015)
- [Data Shapley: Equitable Valuation of Data for Machine Learning](https://mlsanity.com/paper/1904.02868/) (2019)
- [A White Paper on Neural Network Quantization](https://mlsanity.com/paper/2106.08295/) (2021)
- [Assessing Generative Models via Precision and Recall](https://mlsanity.com/paper/1806.00035/) (2018)
- [Universality of Deep Convolutional Neural Networks](https://mlsanity.com/paper/1805.10769/) (2018)
- [Data Banzhaf: A Robust Data Valuation Framework for Machine Learning](https://mlsanity.com/paper/2205.15466/) (2022)
- [We Need to Talk About Random Splits](https://mlsanity.com/paper/2005.00636/) (2020)
- [Randomness In Neural Network Training: Characterizing The Impact of Tooling](https://mlsanity.com/paper/2106.11872/) (2021)
- [Nondeterminism and Instability in Neural Network Optimization](https://mlsanity.com/paper/2103.04514/) (2021)
- [InSituNet: Deep Image Synthesis for Parameter Space Exploration of Ensemble Simulations](https://mlsanity.com/paper/1908.00407/) (2019)
- [GNN-Surrogate: A Hierarchical and Adaptive Graph Neural Network for Parameter Space Exploration of Unstructured-Mesh Ocean Simulations](https://mlsanity.com/paper/2202.08956/) (2022)
- [Convergence Analysis of Distributed Stochastic Gradient Descent with Shuffling](https://mlsanity.com/paper/1709.10432/) (2017)
- [VDL-Surrogate: A View-Dependent Latent-based Model for Parameter Space Exploration of Ensemble Simulations](https://mlsanity.com/paper/2207.13091/) (2022)
- [NNVA: Neural Network Assisted Visual Analysis of Yeast Cell Polarization Simulation](https://mlsanity.com/paper/1904.09044/) (2019)
- [On the Impact of Lossy Image and Video Compression on the Performance of Deep Convolutional Neural Network Architectures](https://mlsanity.com/paper/2007.14314/) (2020)
- [Generalization Through The Lens Of Leave-One-Out Error](https://mlsanity.com/paper/2203.03443/) (2022)
- [COMET: A Novel Memory-Efficient Deep Learning Training Framework by Using Error-Bounded Lossy Compression](https://mlsanity.com/paper/2111.09562/) (2021)
- [Understanding The Effectiveness of Lossy Compression in Machine Learning Training Sets](https://mlsanity.com/paper/2403.15953/) (2024)
- [Machine Learning Visualization Tool for Exploring Parameterized Hydrodynamics](https://mlsanity.com/paper/2406.15509/) (2024)

## Built on by (most cited first, 1 of 1)

- [Transmission Factors for Lossy Compression of PDE Training Data: Measuring What Reaches a Trained Operator](https://mlsanity.com/paper/2610.06095/) (2026)

## Abstract (arXiv)

Neural networks are used as generative surrogate models for scientific discovery, which are trainable approximations of scientific simulations. These models enable users to replace time-consuming numerical simulations with learned alternatives, providing quick solutions. However, high-fidelity generative surrogate models require massive training datasets, which can create storage and I/O challenges. Lossy compression is a promising way to reduce this burden, but compression errors may affect the model quality in subtle ways, making it challenging to quantify their impact. In this work, we examine how lossy compression of training data impacts the quality of generative surrogate models. We begin by characterizing the uncertainty inherent in training neural networks, showing that identical training configurations can produce different models. By exploiting this variability, we propose a method to estimate how much compression-induced error a surrogate model can tolerate without affecting its accuracy. Evaluation of two application simulations demonstrates that our approach significantly reduces memory/storage requirements and speeds up training while producing high-quality surrogate models. These results show that lossy compression saves data storage up to 23.7x and 39x with negligible impact on the quality of the surrogate model. Meanwhile, reducing the size of the training data set also enhances the data loading speed and reduces the training time by up to 3x.