Research map / Image generation / Fast diffusion sampling
Quantifying Error Propagation and Model Collapse in Diffusion Models
Nail B. Khelifa, Richard E. Turner, Ramji Venkataramanan
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 fresh data fraction alpha from 0.1 to 0.9 in recursive training (ablation) | recursively trained diffusion model · 10-dim Gaussian mixture, 20 generations | drift from target distribution | Improved |
| use moderate fresh data fraction alpha = 0.5 (ablation) | recursively trained diffusion model · 10-dim Gaussian mixture, 20 generations | drift from target distribution | Mixed |
| increase fresh data fraction alpha (ablation) | recursive training, error memory · Gaussian mixture, accumulated divergence decomposition | contribution of past score errors | Improved |
| use state-dependent score perturbations instead of time-only perturbations (ablation) | CIFAR-10 diffusion sampling · observability coefficient estimate | observability eta | Improved |
| score error energy persists above a floor across generations | recursive diffusion training (theory) · accumulated chi-squared divergence | accumulated divergence D_N | Hurt |
Closest papers by meaning
- Evaluating the design space of diffusion-based generative models (NeurIPS 2024 · similarity 0.90)
- The Emergence of Reproducibility and Consistency in Diffusion Models (ICML 2024 · similarity 0.90)
- Neural Network-Based Score Estimation in Diffusion Models: Optimization and Generalization (ICLR 2024 · similarity 0.90)
- Algorithm- and Data-Dependent Generalization Bounds for Diffusion Models (NeurIPS 2025 · similarity 0.89)
- O(d/T) Convergence Theory for Diffusion Probabilistic Models under Minimal Assumptions (ICLR 2025 · similarity 0.89)
- Accelerating Convergence of Score-Based Diffusion Models, Provably (ICML 2024 · similarity 0.89)
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.
