Research map / Graph learning / Knowledge and temporal graphs
Learning Dynamic Graph Embeddings With Neural Controlled Differential Equations
Tiexin Qin, Benjamin Walker, Terry Lyons, Hong Yan, Haoliang Li
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 natural cubic instead of rectilinear interpolation (ablation) | GN-CDE · Heat Diffusion (Grid) | MAE: 0.873 -> 0.369 | Improved |
| use natural cubic instead of linear interpolation (ablation) | GN-CDE · Heat Diffusion (Grid) | MAE: 0.453 -> 0.369 | Improved |
| use natural cubic instead of cubic Hermite interpolation (ablation) | GN-CDE · Heat Diffusion (Random) | MAE: 0.523 -> 0.521 | Improved |
| put the graph adjacency into the CDE vector field (GN-CDE) instead of plain Neural CDE | Neural CDE · Heat Diffusion (Community) | MAE: 1.857 -> 0.457 | Improved |
| put the graph adjacency into the CDE vector field (GN-CDE) instead of plain Neural CDE | Neural CDE · Gene Regulation (Random) | MAE: 6.107 -> 2.193 | Improved |
| replace STG-NCDE by GN-CDE | STG-NCDE · Gene Regulation (Grid) | MAE: 6.554 -> 1.388 | Improved |
| use continuous graph path control instead of Neural ODE | Neural ODE · Heat Diffusion (Random) | MAE: 0.629 -> 0.521 | Improved |
| use GN-CDE instead of Neural CDE on real data | Neural CDE · tgbn-genre | NDCG@10: 0.478 -> 0.529 | Improved |
| use GN-CDE instead of Neural ODE on real data | Neural ODE · tgbn-trade | NDCG@10: 0.725 -> 0.731 | No clear effect |
| use GN-CDE instead of task-tailored TGNv2 | TGNv2 · tgbn-trade | NDCG@10: 0.736 -> 0.731 | Hurt |
| use GN-CDE instead of DyGFormer | DyGFormer · tgbn-genre | NDCG@10: 0.365 -> 0.529 | Improved |
Closest papers by meaning
- Permutation Equivariant Neural Controlled Differential Equations for Dynamic Graph Representation Learning (NeurIPS 2025 · similarity 0.92)
- Signed Graph Neural Ordinary Differential Equation for Modeling Continuous-Time Dynamics (AAAI 2024 · similarity 0.91)
- CSG-ODE: ControlSynth Graph ODE For Modeling Complex Evolution of Dynamic Graphs (ICML 2025 · similarity 0.90)
- When Graph Neural Networks Meet Dynamic Mode Decomposition (ICLR 2025 · similarity 0.89)
- A Deep Probabilistic Framework for Continuous Time Dynamic Graph Generation (AAAI 2025 · similarity 0.88)
- State Space Models on Temporal Graphs: A First-Principles Study (NeurIPS 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.
