Research map / Architectures and efficiency / Pruning and efficient tuning
An Operator Theoretic View On Pruning Deep Neural Networks
William T Redman, MARIA FONOBEROVA, Ryan Mohr, Yannis Kevrekidis, Igor Mezic
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Use Koopman magnitude pruning instead of global magnitude pruning | MnistNet · MNIST | accuracy | No clear effect |
| Use Koopman magnitude pruning instead of global magnitude pruning | ResNet-20 · CIFAR-10 | accuracy | No clear effect |
| Use Koopman magnitude pruning instead of global magnitude pruning | MnistNet · MNIST, epochs 1 and 20 | accuracy | No clear effect |
| Use Koopman magnitude pruning instead of global magnitude pruning | ResNet-20 · CIFAR-10, epochs 1 and 20 | accuracy | No clear effect |
| Use global magnitude pruning instead of layer shuffle pruning | ResNet-20 · CIFAR-10, epoch 1 | accuracy | Hurt |
| Use Koopman magnitude pruning instead of layer shuffle pruning | ResNet-20 · CIFAR-10, epoch 1 | accuracy | Hurt |
| Use global magnitude pruning instead of layer shuffle pruning | ResNet-20 · CIFAR-10, epoch 20, compression below 16 | accuracy | Improved |
| Use Koopman magnitude pruning instead of layer shuffle pruning | ResNet-20 · CIFAR-10, epoch 20, compression below 16 | accuracy | Improved |
| Use global magnitude pruning instead of layer shuffle pruning | MnistNet · MNIST, epoch 1, compression above 8 | accuracy | Improved |
| Use Koopman magnitude pruning instead of layer shuffle pruning | MnistNet · MNIST, epoch 1, compression above 8 | accuracy | Improved |
| Use Koopman gradient pruning instead of just gradient pruning | ResNet-20 · CIFAR-10, epoch 20 | accuracy | No clear effect |
Closest papers by meaning
- Pruning via Sparsity-indexed ODE: a Continuous Sparsity Viewpoint (ICML 2023 · similarity 0.90)
- Optimizing Neural Networks via Koopman Operator Theory (NeurIPS 2020 · similarity 0.90)
- Pruning Neural Networks at Initialization: Why Are We Missing the Mark? (ICLR 2021 · similarity 0.90)
- Pruning neural networks without any data by iteratively conserving synaptic flow (NeurIPS 2020 · similarity 0.90)
- Lookahead: A Far-sighted Alternative of Magnitude-based Pruning (ICLR 2020 · similarity 0.90)
- How Sparse Can We Prune A Deep Network: A Fundamental Limit Perspective (NeurIPS 2024 · similarity 0.88)
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.
