Research map / Probabilistic and scientific modeling / Variational and generative sampling
Functional Mean Flow in Hilbert Space
Zhiqi Li, Yuchen Sun, Greg Turk, Bo Zhu
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 |
|---|---|---|---|
| use x1-prediction instead of u-prediction (ablation) | Functional Mean Flow · Navier-Stokes | spectrum MSE: 1.2e3 -> 5.6e2 | Improved |
| use x1-prediction instead of u-prediction (ablation) | Functional Mean Flow · Navier-Stokes | density MSE: 9.7e-5 -> 8.0e-5 | Improved |
| use x1-prediction instead of u-prediction (ablation) | Functional Mean Flow · 1D statistical datasets | statistical MSEs: Genes variance 3.3e-4 -> 2.0e-3; Pop. autocorrelation 7.2e-3 -> 1.1e-4 | Mixed |
| use x1-prediction instead of u-prediction (ablation) | Functional Mean Flow on MNIST SDF (Functional Diffusion framework) · MNIST SDF training stability | training stability (spatial variance collapse) | Improved |
| replace multi-step functional models with one-step FMF | Infty-Diff image generation · CelebA-HQ 64 | FID_CLIP: 4.57 -> 3.48 | Improved |
| replace multi-step functional models with one-step FMF | Infty-Diff image generation · LSUN-Church-256 | FID_CLIP: 10.36 -> 26.57 | Hurt |
| use FMF instead of other one-step functional generators | one-step functional image generation · FFHQ-256 | FID_CLIP: 24.37 -> 11.37 | Improved |
| generate at higher resolution than the training grid (ablation) | FMF on CelebA-HQ (trained on 1/4 of 256x256 pixels) · CelebA-HQ | FID_CLIP: 3.48 (64) -> 10.96 (1024) | Hurt |
| use one-step FMF instead of 64-step Functional Diffusion | ShapeNet 64-point reconstruction · ShapeNet | Chamfer: 0.101 -> 0.060 Chamfer; F-Score 0.707 -> 0.584 | Mixed |
| use FMF instead of GANO (one-step) | Navier-Stokes · Navier-Stokes | spectrum MSE: 3.2e4 -> 5.6e2 | Improved |
Closest papers by meaning
- Improving Function Space Flow Matching with Kernel Optimal Transport (NeurIPS 2026 · similarity 0.89)
- Mean Flows for One-step Generative Modeling (NeurIPS 2025 · similarity 0.89)
- An Error Analysis of Flow Matching for Deep Generative Modeling (ICML 2025 · similarity 0.87)
- Optimal Flow Matching: Learning Straight Trajectories in Just One Step (NeurIPS 2024 · similarity 0.86)
- Flow matching achieves almost minimax optimal convergence (ICLR 2025 · similarity 0.86)
- Structured Coupling for Flow Matching (NeurIPS 2026 · 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.
