Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Go with the flow: Adaptive control for Neural ODEs
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 open-loop data-conditioned control instead of ANODE | N-CODE image classifier · MNIST | accuracy: 98.2% -> 99.2% | Helped |
| Use open-loop data-conditioned control instead of NODE DC | N-CODE image classifier · MNIST | accuracy: 98.5% -> 99.2% | Helped |
| Use open-loop data-conditioned control instead of ANODE | N-CODE image classifier · CIFAR-10 | accuracy: 61.2% -> 74.1% | Helped |
| Use open-loop data-conditioned control instead of NODE DC | N-CODE image classifier · CIFAR-10 | accuracy: 62.0% -> 74.1% | Helped |
| Add data-dependent latent flows to the autoencoder | AutoN-CODE vs auto-encoder (100 comp.) · CIFAR-10 | FID: 73.24 -> 34.90 | Helped |
| Add data-dependent latent flows to the autoencoder | AutoN-CODE vs auto-encoder (1000 comp.) · CIFAR-10 | FID: 55.55 -> 24.19 | Helped |
| Add data-dependent latent flows to the autoencoder | AutoN-CODE vs auto-encoder (100 comp.) · CelebA | FID: 63.11 -> 55.27 | Helped |
| Add data-dependent latent flows to the autoencoder | AutoN-CODE vs auto-encoder (1000 comp.) · CelebA | FID: 55.60 -> 52.47 | Helped |
| Increase the latent Gaussian-mixture components from 100 to 1000 (ablation) | AutoN-CODE · CIFAR-10 | FID: 34.90 -> 24.19 | Helped |
| Increase the latent Gaussian-mixture components from 100 to 1000 (ablation) | AutoN-CODE · CelebA | FID: 55.27 -> 52.47 | Helped |
| Use closed-loop dynamic synapses for sequential pattern memorization | N-CODE · Few-shot pattern memorization | reconstruction error: below 1% | Helped |
Closest papers by meaning
- How Deep Do We Need: Accelerating Training and Inference of Neural ODEs via Control Perspective (ICML 2024 · similarity 0.91)
- ControlSynth Neural ODEs: Modeling Dynamical Systems with Guaranteed Convergence (NeurIPS 2024 · similarity 0.89)
- Simulation-Free Training of Neural ODEs on Paired Data (NeurIPS 2024 · similarity 0.89)
- Neural Context Flows for Meta-Learning of Dynamical Systems (ICLR 2025 · similarity 0.88)
- Neural Conjugate Flows: A Physics-Informed Architecture with Flow Structure (AAAI 2025 · similarity 0.88)
- Semi-supervised Learning of Dynamical Systems with Neural Ordinary Differential Equations: A Teacher-Student Model Approach (AAAI 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.