Research map / Probabilistic and scientific modeling / Differential equations and PDEs
On the Validity of Modeling SGD with Stochastic Differential Equations (SDEs)
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 SGD noise with Gaussian noise of matching covariance (NGD) (ablation) | SGD batch 125 · CIFAR-10 | test accuracy: 89.4 -> 89.3 | No clear effect |
| add SGD noise (SGD vs full-batch GD) | GD · CIFAR-10 | test accuracy: 75.5 -> 89.4 | Helped |
| interpolate SGD toward SDE via SVAG (l>1) in standard settings (ablation) | SGD · CIFAR-10 vision tasks | test accuracy | No clear effect |
| increase batch size to 1024 under linear scaling rule | SGD PreResNet32 · CIFAR-10 | test accuracy | Hurt |
Closest papers by meaning
- On the SDEs and Scaling Rules for Adaptive Gradient Algorithms (NeurIPS 2022 · similarity 0.87)
- High-dimensional Limit of SGD for Diagonal Linear Networks (NeurIPS 2026 · similarity 0.86)
- Efficient Training of Neural Stochastic Differential Equations by Matching Finite Dimensional Distributions (ICLR 2025 · similarity 0.86)
- Neural Stochastic Flows: Solver-Free Modelling and Inference for SDE Solutions (NeurIPS 2025 · similarity 0.85)
- Emergence and scaling laws in SGD learning of shallow neural networks (NeurIPS 2025 · similarity 0.85)
- Integration Matters for Learning PDEs with Backwards SDEs (NeurIPS 2025 · similarity 0.85)
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.