mlsanityData to 7 Oct 2026

Machine-Learning Non-Conservative Dynamics for New-Physics Detection

arXiv 2106.00026 ↗First posted 31 May 2021cs.LG astro-ph.IM gr-qccited by 10

Not on the map. Why?

Abstract (arXiv)

Energy conservation is a basic physics principle, the breakdown of which often implies new physics. This paper presents a method for data-driven "new physics" discovery. Specifically, given a trajectory governed by unknown forces, our Neural New-Physics Detector (NNPhD) aims to detect new physics by decomposing the force field into conservative and non-conservative components, which are represented by a Lagrangian Neural Network (LNN) and a universal approximator network (UAN), respectively, trained to minimize the force recovery error plus a constant λ times the magnitude of the predicted non-conservative force. We show that a phase transition occurs at λ=1, universally for arbitrary forces. We demonstrate that NNPhD successfully discovers new physics in toy numerical experiments, rediscovering friction (1493) from a damped double pendulum, Neptune from Uranus' orbit (1846) and gravitational waves (2017) from an inspiraling orbit. We also show how NNPhD coupled with an integrator outperforms previous methods for predicting the future of a damped double pendulum.

Key facts. Cited by 10 papers in the map (data through 7 Oct 2026). Builds on 19 of its 41 references that are in the map. First posted on arXiv 31 May 2021.

Builds on

19 of 41 references.

References · 19

  1. 2019Hamiltonian Neural Networkscited by 702
  2. 2016Interaction Networks for Learning about Objects, Relations and Physicscited by 503
  3. 2020Lagrangian Neural Networkscited by 388
  4. 2020Discovering Symbolic Models from Deep Learning with Inductive Biasescited by 335
  5. 2017PDE-Net: Learning PDEs from Datacited by 304
  6. 2019Deep Lagrangian Networks: Using Physics as Model Prior for Deep Learningcited by 205
  7. 2019Hamiltonian Generative Networkscited by 136
  8. 2020Simplifying Hamiltonian and Lagrangian Neural Networks via Explicit Constraintscited by 102
  9. 2016Why does deep and cheap learning work so well?cited by 100
  10. 2019Integration of Neural Network-Based Symbolic Regression in Deep Learning for Scientific Discoverycited by 93
  11. 2020Augmenting Physical Models with Deep Networks for Complex Dynamics Forecastingcited by 93
  12. 2020Dissipative SymODEN: Encoding Hamiltonian Dynamics with Dissipation and Control into Deep Learningcited by 77
7 more
  1. 2020AI Poincar'e: Machine Learning Conservation Laws from Trajectoriescited by 77
  2. 2018Augmenting Physical Simulators with Stochastic Neural Networks: Case Study of Planar Pushing and Bouncingcited by 36
  3. 2019Extracting Interpretable Physical Parameters from Spatiotemporal Systems using Unsupervised Learningcited by 23
  4. 2020Symbolic Pregression: Discovering Physical Laws from Distorted Videocited by 21
  5. 2020Forecasting Hamiltonian dynamics without canonical coordinatescited by 17
  6. 2019Neural Canonical Transformation with Symplectic Flowscited by 16
  7. 2018Toward an AI Physicist for Unsupervised Learningcited by 12

Built on by

Cited by year

Citing papers per year: 2021: 4; 2022: 3; 2023: 1; 2024: 2.

10 papers.

  1. 2021SE(3) Equivariant Graph Neural Networks with Complete Local Framescited by 75
  2. 2022Rethinking Symbolic Regression Datasets and Benchmarks for Scientific Discoverycited by 44
  3. 2021Machine-learning hidden symmetriescited by 31
  4. 2022Deconstructing the Inductive Biases of Hamiltonian Neural Networkscited by 31
  5. 2022Multiscale Neural Operator: Learning Fast and Grid-independent PDE Solverscited by 12
  6. 2021Physics-Augmented Learning: A New Paradigm Beyond Physics-Informed Learningcited by 11
  7. 2024A Triumvirate of AI Driven Theoretical Discoverycited by 5
  8. 2021Extracting Dynamical Models from Datacited by 1
  9. 2023Discovering Symbolic Laws Directly from Trajectories with Hamiltonian Graph Neural Networkscited by 1
  10. 2024A Phenomenological AI Foundation Model for Physical Signalscited by 1