Progressive Magnitude-Based Pruning Achieves Sparse Neural Networks in Single Training Cycle
Researchers propose a single-cycle neural network pruning method that gradually increases sparsity during training using a linear schedule, achieving competitive or superior accuracy compared to multi-cycle iterative approaches. The method was benchmarked against established baselines—LTH, SNIP, and GraSP—on CIFAR-10 and MNIST across ResNet, VGG-style, and LeNet architectures. The findings suggest that computationally expensive iterative pruning cycles may not be necessary to obtain high-quality sparse subnetworks.
A preprint submitted to arXiv presents progressive magnitude-based pruning as a practical alternative to iterative pruning methods such as the Lottery Ticket Hypothesis (LTH), which require multiple full training cycles. The proposed approach incrementally raises sparsity throughout a single training run via a linear schedule, updating pruning masks based on the magnitudes of currently active weights. On CIFAR-10, the method achieves 95.12% accuracy on ResNet-18 at 72.9% sparsity, outperforming LTH's reported 90.5%. At extreme sparsity levels, it reaches 93.13% on a VGG-like architecture at 97% sparsity versus SNIP's ~92.0%, and 93.44% on VGG-19 at 97.97% sparsity versus GraSP's 92.19% at 98% sparsity. A sparsity-accuracy analysis on ResNet-18 further shows accuracy remains within 0.1 percentage points of the dense baseline across the 70–85% sparsity range. The authors conclude that single-cycle progressive pruning offers an efficient and effective path to neural network sparsification without sacrificing predictive performance.
What's missing
The study evaluates only image classification tasks on relatively small datasets (CIFAR-10, MNIST); generalization to large-scale datasets (e.g., ImageNet), natural language processing, or other modalities is not demonstrated. Comparisons with baselines rely on previously reported figures rather than controlled re-implementations, which may introduce inconsistencies in training conditions. Wall-clock training time and hardware efficiency gains from the reduced sparsification overhead are not quantified. The work has not yet undergone peer review.
What different sources said
- arXiv cs.LGCenter
Finding Sparse Subnetworks in One Training Cycle via Progressive Magnitude-Based Pruning
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.