Research map / Probabilistic and scientific modeling / Dynamical systems
Constants of motion network
Muhammad Firmansyah Kasim, Yi Heng Lim
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 COMET instead of NODE | dynamics model · damped pendulum | RMSE: 0.14 -> 0.007 | Improved |
| use COMET instead of HNN on non-canonical coordinates | dynamics model · damped pendulum | RMSE: 110 -> 0.007 | Improved |
| use COMET instead of NODE | dynamics model · two body | RMSE: 460 -> 0.42 | Improved |
| use COMET instead of NODE | dynamics model · Lotka-Volterra | RMSE: 0.12 -> 0.048 | Improved |
| use COMET instead of NODE | dynamics model · nonlinear spring | RMSE: 0.63 -> 0.23 | Improved |
| use COMET instead of LNN | dynamics model · 2D pendulum | RMSE: 0.029 -> 0.18 | Hurt |
| use COMET instead of LNN | dynamics model · mass-spring | RMSE: 0.12 -> 0.10 | No clear effect |
| increase constants of motion from 1 to 2 (ablation) | COMET · damped pendulum | RMSE: 0.019 -> 0.007 | Improved |
| increase constants of motion from 1 to 2 (ablation) | COMET · two body | RMSE: 0.79 -> 0.64 | Improved |
| increase constants of motion from 1 to 2 (ablation) | COMET · nonlinear spring | RMSE: 0.37 -> 0.23 | Improved |
| increase constants of motion from 2 to full (3) (ablation) | COMET · 2D pendulum | RMSE: 1.3 -> 0.18 | Improved |
| increase constants of motion from 2 to full (7) (ablation) | COMET · two body | RMSE: 0.64 -> 0.42 | Improved |
Closest papers by meaning
- Liquid Time-constant Networks (AAAI 2021 · similarity 0.84)
- Learning Physics From Video: Unsupervised Physical Parameter Estimation for Continuous Dynamical Systems (CVPR 2025 · similarity 0.83)
- Discovering Physics Laws of Dynamical Systems via Invariant Function Learning (ICML 2025 · similarity 0.81)
- Learning Physics Informed Neural ODEs with Partial Measurements (AAAI 2025 · similarity 0.81)
- When are dynamical systems learned from time series data statistically accurate? (NeurIPS 2024 · similarity 0.81)
- Towards Cross Domain Generalization of Hamiltonian Representation via Meta Learning (ICLR 2024 · similarity 0.81)
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.
