# Continual Learning with Dynamic Sparse Training: Exploring Algorithms for Effective Model Updates

arXiv: https://arxiv.org/abs/2308.14831 · HTML: https://mlsanity.com/paper/2308.14831/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 34 of its 46 references that are in the map.
- First posted on arXiv 28 Aug 2023.

## Builds on (most cited first, 30 of 34)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://mlsanity.com/paper/1801.04381/) (2018)
- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [Learning without Forgetting](https://mlsanity.com/paper/1606.09282/) (2016)
- [iCaRL: Incremental Classifier and Representation Learning](https://mlsanity.com/paper/1611.07725/) (2016)
- [Matching Networks for One Shot Learning](https://mlsanity.com/paper/1606.04080/) (2016)
- [Gradient Episodic Memory for Continual Learning](https://mlsanity.com/paper/1706.08840/) (2017)
- [Continual Learning with Deep Generative Replay](https://mlsanity.com/paper/1705.08690/) (2017)
- [PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning](https://mlsanity.com/paper/1711.05769/) (2017)
- [Sparsity in Deep Learning: Pruning and growth for efficient inference and training in neural networks](https://mlsanity.com/paper/2102.00554/) (2021)
- [Rigging the Lottery: Making All Tickets Winners](https://mlsanity.com/paper/1911.11134/) (2019)
- [Expert Gate: Lifelong Learning with a Network of Experts](https://mlsanity.com/paper/1611.06194/) (2016)
- [DER: Dynamically Expandable Representation for Class Incremental Learning](https://mlsanity.com/paper/2103.16788/) (2021)
- [Scalable Training of Artificial Neural Networks with Adaptive Sparse Connectivity inspired by Network Science](https://mlsanity.com/paper/1707.04780/) (2017)
- [FOSTER: Feature Boosting and Compression for Class-Incremental Learning](https://mlsanity.com/paper/2204.04662/) (2022)
- [Supermasks in Superposition](https://mlsanity.com/paper/2006.14769/) (2020)
- [Sparse Networks from Scratch: Faster Training without Losing Performance](https://mlsanity.com/paper/1907.04840/) (2019)
- [A Model or 603 Exemplars: Towards Memory-Efficient Class-Incremental Learning](https://mlsanity.com/paper/2205.13218/) (2022)
- [Don't forget, there is more than forgetting: new metrics for Continual Learning](https://mlsanity.com/paper/1810.13166/) (2018)
- [Class-Incremental Learning by Knowledge Distillation with Adaptive Feature Consolidation](https://mlsanity.com/paper/2204.00895/) (2022)
- [Do We Actually Need Dense Over-Parameterization? In-Time Over-Parameterization in Sparse Training](https://mlsanity.com/paper/2102.02887/) (2021)
- [Continual Learning via Neural Pruning](https://mlsanity.com/paper/1903.04476/) (2019)
- [Top-KAST: Top-K Always Sparse Training](https://mlsanity.com/paper/2106.03517/) (2021)
- [RMM: Reinforced Memory Management for Class-Incremental Learning](https://mlsanity.com/paper/2301.05792/) (2023)
- [SparCL: Sparse Continual Learning on the Edge](https://mlsanity.com/paper/2209.09476/) (2022)
- [SpaceNet: Make Free Space For Continual Learning](https://mlsanity.com/paper/2007.07617/) (2020)
- [NISPA: Neuro-Inspired Stability-Plasticity Adaptation for Continual Learning in Sparse Networks](https://mlsanity.com/paper/2206.09117/) (2022)
- [Ten Lessons We Have Learned in the New "Sparseland": A Short Handbook for Sparse Neural Network Researchers](https://mlsanity.com/paper/2302.02596/) (2023)

## Built on by (most cited first, 8 of 8)

- [Self-Regulated Neurogenesis for Online Data-Incremental Learning](https://mlsanity.com/paper/2403.14684/) (2024)
- [More Experts Than Galaxies: Conditionally-overlapping Experts With Biologically-Inspired Fixed Routing](https://mlsanity.com/paper/2410.08003/) (2024)
- [Similarity-based context aware continual learning for spiking neural networks](https://mlsanity.com/paper/2411.05802/) (2024)
- [Continual Learning of Multiple Cognitive Functions with Brain-inspired Temporal Development Mechanism](https://mlsanity.com/paper/2504.05621/) (2025)
- [Unlocking [CLS] Features for Continual Post-Training](https://mlsanity.com/paper/2502.14762/) (2025)
- [NeuroTrails: Training with Dynamic Sparse Heads as the Key to Effective Ensembling](https://mlsanity.com/paper/2505.17909/) (2025)
- [Pruned Adaptation Modules: A Simple yet Strong Baseline for Continual Foundation Models](https://mlsanity.com/paper/2603.21170/) (2026)
- [CLARE: Scalable Class-Incremental Continual Learning via a Sparsity-Based Framework](https://mlsanity.com/paper/2609.17026/) (2026)

## Abstract (arXiv)

Continual learning (CL) refers to the ability of an intelligent system to sequentially acquire and retain knowledge from a stream of data with as little computational overhead as possible. To this end; regularization, replay, architecture, and parameter isolation approaches were introduced to the literature. Parameter isolation using a sparse network which enables to allocate distinct parts of the neural network to different tasks and also allows to share of parameters between tasks if they are similar. Dynamic Sparse Training (DST) is a prominent way to find these sparse networks and isolate them for each task. This paper is the first empirical study investigating the effect of different DST components under the CL paradigm to fill a critical research gap and shed light on the optimal configuration of DST for CL if it exists. Therefore, we perform a comprehensive study in which we investigate various DST components to find the best topology per task on well-known CIFAR100 and miniImageNet benchmarks in a task-incremental CL setup since our primary focus is to evaluate the performance of various DST criteria, rather than the process of mask selection. We found that, at a low sparsity level, Erdos-R'enyi Kernel (ERK) initialization utilizes the backbone more efficiently and allows to effectively learn increments of tasks. At a high sparsity level, unless it is extreme, uniform initialization demonstrates a more reliable and robust performance. In terms of growth strategy; performance is dependent on the defined initialization strategy and the extent of sparsity. Finally, adaptivity within DST components is a promising way for better continual learners.