Research map / Image generation / Fast diffusion sampling
Mitigating the Contractivity Trap in Diffusion ODEs via Stein Stabilization
Shigui Li, Delu zeng
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 |
|---|---|---|---|
| add SteinDiff Stein correction to DPM-Solver++ sampling at 5 NFE | DPM-Solver++ (3m) · LSUN-Bedrooms 256x256 (Latent Diffusion) | FID: 18.61 -> 7.64 | Improved |
| add SteinDiff Stein correction to DPM-Solver++ sampling at 6 NFE | DPM-Solver++ (3m) · LSUN-Bedrooms 256x256 (Latent Diffusion) | FID: 8.52 -> 4.71 | Improved |
| add SteinDiff Stein correction to DPM-Solver++ sampling at 8 NFE | DPM-Solver++ (3m) · LSUN-Bedrooms 256x256 (Latent Diffusion) | FID: 4.15 -> 3.72 | Improved |
| add SteinDiff Stein correction to DPM-Solver++ sampling at 10 NFE | DPM-Solver++ (3m) · LSUN-Bedrooms 256x256 (Latent Diffusion) | FID: 3.61 -> 3.38 | Improved |
| add SteinDiff Stein correction to DPM-Solver++ sampling at 12 NFE | DPM-Solver++ (3m) · LSUN-Bedrooms 256x256 (Latent Diffusion) | FID: 3.43 -> 3.01 | Improved |
| add SteinDiff Stein correction to DPM-Solver++ sampling at 15 NFE | DPM-Solver++ (3m) · LSUN-Bedrooms 256x256 (Latent Diffusion) | FID: 3.28 -> 2.86 | Improved |
| add SteinDiff Stein correction to DPM-Solver++ sampling at 20 NFE | DPM-Solver++ (3m) · LSUN-Bedrooms 256x256 (Latent Diffusion) | FID: 3.17 -> 2.77 | Improved |
Closest papers by meaning
- Tortoise and Hare Guidance: Accelerating Diffusion Model Inference with Multirate Integration (NeurIPS 2025 · similarity 0.89)
- Adaptive Stochastic Coefficients for Accelerating Diffusion Sampling (NeurIPS 2025 · similarity 0.88)
- SADA: Stability-guided Adaptive Diffusion Acceleration (ICML 2025 · similarity 0.88)
- EVODiff: Entropy-aware Variance Optimized Diffusion Inference (NeurIPS 2025 · similarity 0.87)
- Self-Refining Diffusion Samplers: Enabling Parallelization via Parareal Iterations (NeurIPS 2024 · similarity 0.87)
- Closing the Approximation Gap in Simulation-free Latent SDEs (NeurIPS 2026 · similarity 0.87)
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.
