Research map / Image generation / Fast diffusion sampling
On the Generalization Properties of Diffusion Models
Puheng Li, Zhong Li, Huishuai Zhang, Jiang Bian
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 |
|---|---|---|---|
| train beyond early-stopping time | diffusion model on Gaussian mixture · 1D 2-mode Gaussian mixture, distance 6 | KL divergence | Hurt |
| increase distance between modes from 6 to 30 | diffusion model · 1D 2-mode Gaussian mixture | generalization (density estimation): 6 -> 30 | Hurt |
| train on farthest clusters instead of nearest clusters | U-Net diffusion model · MNIST K-means clusters | training loss and sample quality | Hurt |
Closest papers by meaning
- Generalization Dynamics of Linear Diffusion Models (NeurIPS 2026 · similarity 0.94)
- Understanding diffusion models requires rethinking (again) generalization (NeurIPS 2026 · similarity 0.93)
- Generalization through variance: how noise shapes inductive biases in diffusion models (ICLR 2025 · similarity 0.92)
- On the Edge of Memorization in Diffusion Models (NeurIPS 2025 · similarity 0.91)
- Understanding Generalizability of Diffusion Models Requires Rethinking the Hidden Gaussian Structure (NeurIPS 2024 · similarity 0.91)
- The Emergence of Reproducibility and Consistency in Diffusion Models (ICML 2024 · similarity 0.91)
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.
