Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Foundation Inference Models for Ordinary Differential Equations
Johannes R. Hübers, Maximilian Mauel, David Berghaus, Patrick Seifner, Ramses J Sanchez
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 FIM-ODE instead of ODEFormer | zero-shot ODE inference · ODEBench trajectory reconstruction | fraction R2>0.9 (%): 63.1% -> 84.4% | Improved |
| use FIM-ODE instead of ODEFormer | zero-shot ODE inference · ODEBench trajectory generalization | fraction R2>0.9 (%): 27.9% -> 29.5% | No clear effect |
| add 49 extra context trajectories to FIM-ODE (ablation) | FIM-ODE zero-shot · VDP forecasting Task 1 | MSE: 0.8(7) -> 0.35(8) | Improved |
| add 49 extra context trajectories to FIM-ODE (ablation) | FIM-ODE zero-shot · VDP forecasting Task 2 | MSE: 2(1) -> 0.9(2) | Improved |
| add 49 extra context trajectories to FIM-ODE (ablation) | FIM-ODE zero-shot · FHN imputation | MSE: 0.9(8) -> 0.43(6) | Improved |
| finetune FIM-ODE on context data (test-MSE checkpoint selection) (ablation) | FIM-ODE · VDP forecasting Task 1 | MSE: 0.90 -> 0.028 | Improved |
| finetune FIM-ODE on context data (test-MSE checkpoint selection) (ablation) | FIM-ODE · VDP forecasting Task 2 | MSE: 0.43 -> 0.022 | Improved |
| finetune FIM-ODE on context data (test-MSE checkpoint selection) (ablation) | FIM-ODE · FHN imputation | MSE: 0.40 -> 0.029 | Improved |
| select finetuning checkpoint by training loss (top-10 mean) instead of test MSE (ablation) | FIM-ODE finetuning · VDP forecasting Task 1 | MSE: 0.028 -> 0.2(1) | Hurt |
| finetune FIM-ODE on subject data (ablation) | FIM-ODE · CMU MoCap Subject 35 short | test MSE: 655.40 -> 6.92 | Improved |
| finetune FIM-ODE on subject data (ablation) | FIM-ODE · CMU MoCap Subject 39 short | test MSE: 295.29 -> 15.56 | Improved |
| finetune FIM-ODE on subject data (ablation) | FIM-ODE · CMU MoCap Subject 09 short | test MSE: 6.10 -> 7.55 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ODEFormer — ODEFormer: Symbolic Regression of Dynamical Systems with Transformers (ICLR 2024) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- In-Context Learning of Stochastic Differential Equations with Foundation Inference Models (NeurIPS 2025 · similarity 0.89)
- ODEFormer: Symbolic Regression of Dynamical Systems with Transformers (ICLR 2024 · similarity 0.88)
- How Deep Do We Need: Accelerating Training and Inference of Neural ODEs via Control Perspective (ICML 2024 · similarity 0.88)
- Curriculum Multiple Shooting for Robust Training of Neural and Universal Differential Equations (NeurIPS 2026 · similarity 0.87)
- Simulation-Free Training of Neural ODEs on Paired Data (NeurIPS 2024 · similarity 0.86)
- From Fourier to Neural ODEs: Flow Matching for Modeling Complex Systems (ICML 2024 · similarity 0.86)
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.
