Research map / Time series / Foundation models for time series
Sparsifying Networks via Subdifferential Inclusion
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Apply SIS (subdifferential inclusion sparsification) to pre-trained networks | ResNet50 ImageNet · ImageNet 80%/90%/96.5% sparsity | top-1 accuracy | Helped |
| Apply SIS to MobileNetV3 instead of RigL | MobileNet V3 ImageNet · ImageNet 90% sparsity | top-1 accuracy: +0.94 over RigL (66.07 vs 65.13) | Helped |
| Apply SIS to MobileNetV2 instead of RigL | MobileNet V2 ImageNet · ImageNet 90% sparsity | top-1 accuracy | Hurt |
| Apply SIS to VGG19 instead of RigL/LRR on CIFAR-10 | VGG19 CIFAR-10 · CIFAR-10 90%/95% sparsity | test accuracy | Mixed |
| Apply SIS to Transformer-XL instead of RigL/SNIP | Transformer-XL WikiText-103 · 70% sparsity | perplexity: 21.1 vs 22.4 (RigL) | Helped |
| Apply SIS to Jasper instead of LRR | Jasper LibriSpeech · 70% sparsity | WER: 13.1 vs 13.7 (LRR) | Helped |
| Apply SIS to N-BEATS instead of RigL | N-BEATS M4 · 70% sparsity | SMAPE: 9.7 vs 10.2 (RigL) | Helped |
| Apply SIS to ResNet50 CIFAR-100 at 90% sparsity instead of baselines | ResNet50 CIFAR-100 · CIFAR-100 90% sparsity | test accuracy | No clear effect |
| Increase accuracy tolerance eta (higher eta = looser constraint) (ablation) | LeNet-FCN MNIST · MNIST | sparsity | Helped |
Closest papers by meaning
- DEEP NEURAL NETWORK INITIALIZATION WITH SPARSITY INDUCING ACTIVATIONS (ICLR 2024 · similarity 0.89)
- Global Minimizers of $\ell^p$-Regularized Objectives Yield the Sparsest ReLU Neural Networks (NeurIPS 2025 · similarity 0.89)
- Random Sparse Lifts: Construction, Analysis and Convergence of finite sparse networks (ICLR 2024 · similarity 0.87)
- Shaving Weights with Occam's Razor: Bayesian Sparsification for Neural Networks using the Marginal Likelihood (NeurIPS 2024 · similarity 0.87)
- A Single-Step, Sharpness-Aware Minimization is All You Need to Achieve Efficient and Accurate Sparse Training (NeurIPS 2024 · similarity 0.87)
- Position: A Theory of Deep Learning Must Include Compositional Sparsity (ICML 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.