Research map / Architectures and efficiency / RNNs and normalization
The High-Dimensional Geometry of Binary Neural Networks
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 |
|---|---|---|---|
| Binarize high-dimensional vectors | Random vectors drawn from a standard normal distribution · Random high-dimensional vectors | angle: arccos √(2/π) ≈ 37° | No clear effect |
| Remove weight binarization from all convolutional layers (ablation) | Binary CNN · CIFAR-10 | classification performance: 3 percent | Hurt |
| Use continuous weights in the first convolutional layer (ablation) | Binary CNN · CIFAR-10 | classification performance: 3 percent | Helped |
| Use continuous weights in each later convolutional layer individually (ablation) | Binary CNN · CIFAR-10 | classification performance | No clear effect |
| Remove activation binarization (ablation) | Binary CNN · CIFAR-10 | classification performance | Hurt |
Closest papers by meaning
- Training Binary Neural Networks via Gaussian Variational Inference and Low-Rank Semidefinite Programming (NeurIPS 2024 · similarity 0.86)
- G-Net: A Provably Easy Construction of High-Accuracy Random Binary Neural Networks (NeurIPS 2025 · similarity 0.86)
- Geometric Inductive Biases of Deep Networks: The Role of Data and Architecture (ICLR 2025 · similarity 0.85)
- Expand-and-Cluster: Parameter Recovery of Neural Networks (ICML 2024 · similarity 0.85)
- Fast and Slow Gradient Approximation for Binary Neural Network Optimization (AAAI 2025 · similarity 0.85)
- Siegel Neural Networks (NeurIPS 2025 · similarity 0.84)
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.