Research map / Image generation / Fast diffusion sampling
On the Collapse of Generative Paths: A Criterion and Correction for Diffusion Steering
Ziseok Lee, Minyeong Hwang, Wooyeol Lee, Sanghyun Jo, Jihyung Ko, Young Bin Park, Jae-Mun Choi, Eunho Yang, Kyungsu Kim
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 adaptive time-varying exponent correction (ACE) to FKC | Feynman-Kac corrector · Synthetic checkerboard | W1 distance: 1.37 -> 0.20 | Improved |
| add bump correction with B1=10 over no bump (B1=0) (ablation) | ACE · Synthetic checkerboard | W1 distance: 0.78 -> 0.20 | Improved |
| increase bump amplitude B1 from 10 to 50 (ablation) | ACE · Synthetic checkerboard | W1 distance: 0.20 -> 0.66 | Hurt |
| add ACE exponent correction and resampling | FKC · CrossDocked scaffold decoration (omega=1.4) | OSR: 0.40 -> 0.75 | Improved |
| add ACE correction | NR · CrossDocked scaffold decoration (omega=1.4) | OSR: 0.47 -> 0.75 | Improved |
| add ACE correction | FKC · CrossDocked scaffold decoration (omega=1.4) | Vina score (avg): -6.24 -> -7.10 | Improved |
| add ACE correction | FKC · CrossDocked scaffold decoration (omega=1.4) | QED (avg): 0.44 -> 0.53 | Improved |
| add resampling step to ACE-lite (ablation) | ACE-lite · CrossDocked scaffold decoration (omega=1.4) | OSR: 0.66 -> 0.75 | Improved |
| remove resampling (ACE-lite) to cut runtime (ablation) | ACE · CrossDocked scaffold decoration | min per molecule: 0.34 -> 0.19 | Improved |
| compose generalist experts with ACE versus specialized fixed-pose model | ACE · CrossDocked scaffold decoration | OSR: 0.47 -> 0.75 | Improved |
| raise guidance scale omega from 1.1 to 1.4 (ablation) | ACE · CrossDocked scaffold decoration | Vina score (avg): -7.02 -> -7.10 | Improved |
| raise guidance scale omega from 1.1 to 1.4 (ablation) | FKC · CrossDocked scaffold decoration | OSR: 0.52 -> 0.40 | Hurt |
Closest papers by meaning
- A General Framework for Inference-time Scaling and Steering of Diffusion Models (ICML 2025 · similarity 0.90)
- Specificity-Aware Diffusion Steering via Variance-Reduced Sequential Monte Carlo (NeurIPS 2026 · similarity 0.89)
- Flow Density Control: Generative Optimization Beyond Entropy-Regularized Fine-Tuning (NeurIPS 2025 · similarity 0.88)
- Role Bias in Diffusion Models: Diagnosing and Mitigating through Intermediate Decomposition (NeurIPS 2025 · similarity 0.88)
- Flow Matching with General Discrete Paths: A Kinetic-Optimal Perspective (ICLR 2025 · similarity 0.88)
- Particle Guidance: non-I.I.D. Diverse Sampling with Diffusion Models (ICLR 2024 · similarity 0.88)
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.
