Research map / Optimization / Nonconvex and smooth optimization
Dimension-Free Bounds for Low-Precision Training
Zheng Li, Christopher M De Sa
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 |
|---|---|---|---|
| increase model dimension d from small to large | SGD and LP-SGD (8-bit and 6-bit) · Synthetic linear regression | noise ball size | No clear effect |
| increase nonzero entries s (changing sigma_1 and L_1) at fixed d=1024 | LP-SGD vs SGD · Synthetic linear regression | loss gap | Hurt |
| assign exponent bits per Theorem 5 (5 of 20 total bits) | LP-SGD with normal FPQ · Synthetic linear regression | noise ball size | Improved |
| vary exponent bits in 16-bit floating point formats | LP-SGD · MNIST | loss gap | Improved |
Closest papers by meaning
- Refined Error Bounds for Several Learning Algorithms (JMLR 2016 · similarity 0.86)
- SWALP : Stochastic Weight Averaging in Low Precision Training (ICML 2019 · similarity 0.86)
- Tighter CMI-Based Generalization Bounds via Stochastic Projection and Quantization (NeurIPS 2025 · similarity 0.86)
- Scaling Laws for Precision in High-Dimensional Linear Regression (ICML 2026 · similarity 0.86)
- On Generalization Error Bounds of Noisy Gradient Methods for Non-Convex Learning (ICLR 2020 · similarity 0.85)
- Scaling Laws for Floating–Point Quantization Training (ICML 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.
