Research map / Architectures and efficiency / RNNs and normalization
Deep Learning with S-Shaped Rectified Linear Activation Units
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 |
|---|---|---|---|
| replace ReLU with SReLU | NIN · CIFAR-100 no augmentation | error rate: 35.96 -> 31.10 | Helped |
| replace ReLU with SReLU | NIN · CIFAR-10 no augmentation | error rate: 9.67 -> 8.41 | Helped |
| replace ReLU with SReLU | GoogLeNet · ImageNet | error rate: 11.1 -> 9.86 | Helped |
| SReLU instead of APL | NIN · CIFAR-10 with augmentation | error rate: 7.51 -> 6.98 | Helped |
| replace ReLU with SReLU | NIN · MNIST | error rate: 0.47 -> 0.35 | Helped |
| channel-wise instead of channel-shared SReLU (ablation) | NIN + SReLU · CIFAR-10 no augmentation | error rate: 9.01 -> 8.41 | Helped |
| adaptive freezing initialization instead of manual init (ablation) | NIN + SReLU · CIFAR-10 no augmentation | error rate: 8.61 -> 8.41 | Helped |
| adaptive freezing initialization instead of manual init (ablation) | NIN + SReLU · CIFAR-100 no augmentation | error rate: 31.36 -> 31.10 | Helped |
Closest papers by meaning
- Smooth Maximum Unit: Smooth Activation Function for Deep Networks Using Smoothing Maximum Technique (CVPR 2022 · similarity 0.86)
- Understanding Deep Neural Networks with Rectified Linear Units (ICLR 2018 · similarity 0.86)
- Understanding and Improving Convolutional Neural Networks via Concatenated Rectified Linear Units (ICML 2016 · similarity 0.85)
- Gompertz Linear Units: Leveraging Asymmetry for Enhanced Learning Dynamics (NeurIPS 2025 · similarity 0.83)
- ReLUs Are Sufficient for Learning Implicit Neural Representations (ICML 2024 · similarity 0.83)
- Compelling ReLU Networks to Exhibit Exponentially Many Linear Regions at Initialization and During Training (ICML 2025 · similarity 0.83)
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.