Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Delay Flow Matching
Bolin Zhao, Xiaoyu Zhang, Yuting Dong, Xin Lu, Wei Lin, Qunxi 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 |
|---|---|---|---|
| add delay terms to the flow (DFM instead of CFM) | KP-CFM · MNIST translation | FID: 45.020 -> 11.747 | Improved |
| use distinct per-fate initial functions instead of constant ones (ablation) | OT-DFM · Mouse hematopoiesis | W2(F): 0.136 -> 0.095 | Improved |
| use distinct per-fate initial functions instead of constant ones (ablation) | OT-DFM · qPCR iPSC | W2(F): 0.234 -> 0.213 | Improved |
| use DFM(D) instead of OT-CFM | OT-CFM · Mouse hematopoiesis | W2(L): 0.378 -> 0.372 | No clear effect |
| use OT-DFM(C) instead of OT-CFM | OT-CFM · qPCR iPSC | W2(F): 0.226 -> 0.234 | Hurt |
| use DFM instead of CFM with independent coupling | I-CFM · CIFAR-10 (NFE 10) | FID: 108.291 -> 54.064 | Improved |
| use DFM(D) instead of OT-CFM, small NFE | OT-CFM · CIFAR-10 (NFE 10) | FID: 78.165 -> 54.222 | Improved |
| use OT-DFM(D) instead of OT-CFM at adaptive NFE | OT-CFM · CIFAR-10 (Adap.) | FID: 6.162 -> 5.191 | Improved |
| set delay tau to ground truth versus 0.6 in autoregulation (ablation) | KP-DFM(C) · Autoregulation interpolation | W2: 2.82e-2 -> 1.53e-2 | Improved |
| use DFM instead of CFM on spiral DDE | OT-CFM · Spiral DDE interpolation | W2: 3.95 -> 0.20 | Improved |
Closest papers by meaning
- Drift Flow Matching (NeurIPS 2026 · similarity 0.90)
- Switched Flow Matching: Eliminating Singularities via Switching ODEs (ICML 2024 · similarity 0.90)
- Optimal Flow Matching: Learning Straight Trajectories in Just One Step (NeurIPS 2024 · similarity 0.89)
- Beckmann Transport Models: From Autonomous Flows to One-Step Maps (NeurIPS 2026 · similarity 0.89)
- Elucidating Flow Matching ODE Dynamics via Data Geometry and Denoisers (ICML 2025 · similarity 0.89)
- An Error Analysis of Flow Matching for Deep Generative Modeling (ICML 2025 · 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.
