Research map / Architectures and efficiency / RNNs and normalization
Constructive Universal High-Dimensional Distribution Generation through Deep ReLU 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 |
|---|---|---|---|
| Increase the sawtooth depth s | Constructive deep ReLU generator · 2D Lipschitz-continuous target distributions | Wasserstein distance upper bound: \frac{L\sqrt{2}}{2n} + \frac{2\sqrt{2}}{n2^s} | Helped |
| Increase histogram resolution n | Constructive deep ReLU generator · 2D Lipschitz-continuous target distributions | Wasserstein distance upper bound: \frac{L\sqrt{2}}{2n} + \frac{2\sqrt{2}}{n2^s} | Helped |
| Generate from one-dimensional uniform noise rather than d independent random variables | Constructive deep ReLU generator · d-dimensional target distributions | Wasserstein approximation error | No clear effect |
| Use increasing depth in the histogram transport construction | Constructive deep ReLU generator · d-dimensional histogram distributions | Wasserstein distance upper bound: \frac{\sqrt{d}}{n2^s} | Helped |
Closest papers by meaning
- Size-Noise Tradeoffs in Generative Networks (NeurIPS 2018 · similarity 0.89)
- A Universal Approximation Theorem of Deep Neural Networks for Expressing Probability Distributions (NeurIPS 2020 · similarity 0.88)
- Compelling ReLU Networks to Exhibit Exponentially Many Linear Regions at Initialization and During Training (ICML 2025 · similarity 0.86)
- Exact Upper and Lower Bounds for the Output Distribution of Neural Networks with Random Inputs (ICML 2025 · similarity 0.84)
- Characterizing ResNet's Universal Approximation Capability (ICML 2024 · similarity 0.84)
- A Likelihood Based Approach to Distribution Regression Using Conditional Deep Generative Models (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.