Research map / Probabilistic and scientific modeling / Dynamical systems
LieStoNet: Learning Lie Symmetries from Spatiotemporal Data for Stochastic Dynamical Systems
Shida Liu, Abhishek Gupta, Sumit Sinha, L Mahadevan
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 |
|---|---|---|---|
| add Lie-bracket closure loss L1 (ablation) | LieStoNet · synthetic SDE examples | max principal angle | Improved |
| add SDE determining-equation loss L6 (ablation) | LieStoNet · synthetic SDE examples | max principal angle | Improved |
| add finite-step after-push loss L7 (ablation) | LieStoNet · synthetic SDE examples | max principal angle | Improved |
| add each of the remaining algebraic losses (Jacobi, skew-symmetry, bilinearity, independence) (ablation) | LieStoNet · synthetic SDE examples | max principal angle | Improved |
| select generator count m by minimizing closure loss L1 | LieStoNet · four SDE examples | recovered algebra dimension | Improved |
| evaluate at the selected dimension instead of other m | LieStoNet · four SDE examples | max principal angle | Improved |
| push BTC/USDT trajectories along learned generator instead of random direction (eps=0.05) | LieStoNet · BTC/USDT | delta NLL: 9.7 -> 0.008 | Improved |
| push along learned generator 2 instead of random (eps=0.50) | LieStoNet · BTC/USDT | delta NLL: 112.0 -> 17.6 | Improved |
| increase push step length eps from 0.05 to 0.50 for generator 1 | LieStoNet · BTC/USDT | delta NLL: 0.008 -> 6.41 | Hurt |
| increase push step eps from 1 to 5 | LieStoNet · BTC/USDT after-push | drift residual generator 1: 1.4e-3 -> 1.3e-1 | Hurt |
Closest papers by meaning
- Symmetry-Informed Governing Equation Discovery (NeurIPS 2024 · similarity 0.89)
- Learning Infinitesimal Generators of Continuous Symmetries from Data (NeurIPS 2024 · similarity 0.88)
- SymDrift: One-Shot Generative Modeling under Symmetries (NeurIPS 2026 · similarity 0.88)
- Explicit Discovery of Nonlinear Symmetries from Dynamic Data (ICML 2025 · similarity 0.88)
- Latent Mixture of Symmetries for Sample-Efficient Dynamic Learning (NeurIPS 2025 · similarity 0.88)
- Noether's Razor: Learning Conserved Quantities (NeurIPS 2024 · similarity 0.87)
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.
