Research map / Probabilistic and scientific modeling / Variational and generative sampling
Ode to an ODE
Krzysztof M Choromanski, Jared Quincy Davis, Valerii Likhosherstov, Xingyou Song, Jean-Jacques Slotine, Jacob Varley, Honglak Lee, Adrian Weller, Vikas Sindhwani
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 |
|---|---|---|---|
| replace Dense-1 with ODEtoODE-1 | ODEtoODE · Corrupted MNIST Stripe | test accuracy: 30.55 -> 44.51 | Improved |
| replace Dense-1 with ODEtoODE-1 | ODEtoODE · Corrupted MNIST Dotted Lines | test accuracy: 92.99 -> 95.42 | Improved |
| replace Dense-1 with ODEtoODE-1 | ODEtoODE · Corrupted MNIST Glass Blur | test accuracy: 91.62 -> 93.91 | Improved |
| replace HyperNet-10 with ODEtoODE-1 | ODEtoODE · Corrupted MNIST Translate | test accuracy: 29.3 -> 26.82 | Hurt |
| replace HyperNet-10 with ODEtoODE-1 | ODEtoODE · Corrupted MNIST Scale | test accuracy: 68.11 -> 66.68 | Hurt |
| increase number of gates from 1 to 10 (ablation) | ODEtoODE · Corrupted MNIST Stripe | test accuracy: 44.51 -> 44.37 | No clear effect |
| increase number of gates from 1 to 10 (ablation) | ODEtoODE · Corrupted MNIST Motion Blur | test accuracy: 78.63 -> 77.58 | Hurt |
| constant perturbation to hidden units (ODEtoODE-1) works best vs 10 gates (ablation) | ODEtoODE · Corrupted MNIST | test accuracy | Improved |
| add dropout 0.1 and width 256 (Table 2 setting) (ablation) | ODEtoODE-1 · Corrupted MNIST Stripe | test accuracy: 44.51 -> 36.25 | Hurt |
| increase NANODE trigonometric degree beyond 5 (ablation) | NANODE · RL locomotion | reward | Hurt |
| use ODEtoODE policy instead of other architectures | ES policy · Humanoid | reward | Improved |
| orthogonal flow vs HyperNet on ES policies | HyperNet · RL tasks | reward | Hurt |
Closest papers by meaning
- How Deep Do We Need: Accelerating Training and Inference of Neural ODEs via Control Perspective (ICML 2024 · similarity 0.89)
- Simulation-Free Training of Neural ODEs on Paired Data (NeurIPS 2024 · similarity 0.87)
- ControlSynth Neural ODEs: Modeling Dynamical Systems with Guaranteed Convergence (NeurIPS 2024 · similarity 0.87)
- Neural Conjugate Flows: A Physics-Informed Architecture with Flow Structure (AAAI 2025 · similarity 0.86)
- Global Convergence in Neural ODEs: Impact of Activation Functions (ICLR 2025 · similarity 0.86)
- Semi-Implicit Neural Ordinary Differential Equations (AAAI 2025 · 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.
