Hamiltonian Neural Networks
read the original abstract
Even though neural networks enjoy widespread use, they still struggle to learn the basic laws of physics. How might we endow them with better inductive biases? In this paper, we draw inspiration from Hamiltonian mechanics to train models that learn and respect exact conservation laws in an unsupervised manner. We evaluate our models on problems where conservation of energy is important, including the two-body problem and pixel observations of a pendulum. Our model trains faster and generalizes better than a regular neural network. An interesting side effect is that our model is perfectly reversible in time.
This paper has not been read by Pith yet.
Forward citations
Cited by 22 Pith papers
-
When Do Conservation Laws Survive Learned Representations? Certified Horizons for Latent World Models
Shell-horizon certificates are derived for decoded physical invariants in latent models, allowing conservation laws to survive representation learning via a monotone alignment from soft witnesses, with empirical suppo...
-
When Do Conservation Laws Survive Learned Representations? Certified Horizons for Latent World Models
Shell-horizon certificates bound rollout steps on decoded physical invariants from measurable model defects in latent world models, showing some geometric priors survive representation learning while others do not.
-
Between Amnesia and Chaos: A Memory Stability Expressivity Trilemma for Trainable Dissipative Oscillator Networks
Trainable dissipative oscillator networks exhibit a trilemma in which damping governs memory horizon, gradient stability, and Lyapunov exponent, with learned substrates outperforming frozen ones only at short horizons...
-
UWM-JEPA: Predictive World Models That Imagine in Belief Space
UWM-JEPA uses a density-matrix latent and unitary predictor in JEPA to preserve joint-state spectrum during blind rollouts, achieving 0.77 accuracy on a five-step hidden-velocity task versus 0.53 for an LSTM baseline.
-
Hamiltonian Graph Inference Networks: Joint structure discovery and dynamics prediction for lattice Hamiltonian systems from trajectory data
HGIN jointly recovers interaction graphs and predicts trajectories for lattice Hamiltonian systems from data, achieving six to thirteen orders of magnitude lower long-time errors than baselines on Klein-Gordon and dis...
-
Universal Differential Equations for Scientific Machine Learning
Universal Differential Equations unify scientific models with machine learning by embedding flexible approximators into differential equations, enabling applications from biological mechanism discovery to high-dimensi...
-
Chronos: A Physics-Informed Full-History Framework for Non-Markovian Long-Horizon Manipulation
Chronos elevates full observation history to the policy's latent state via selective SSM tokens and a Schrödinger-inspired acceleration bridge, achieving large gains on memory-dependent robot tasks with fewer parameters.
-
Reconstructing Galactic Gravitational Potentials from Stellar Kinematics with Physics-Informed Neural Networks
A PINN approach learns galactic gravitational potentials from acceleration data, achieving sub-percent errors on simulations while outperforming analytic models and retaining interpretability via structured priors.
-
NEXUS: Neural Energy Fields for Physically Consistent Contact-Rich 3D Object Dynamics
NEXUS introduces a graph-based neural energy-field model that derives forces from scalar energy and dissipation terms to achieve physically consistent contact-rich 3D dynamics.
-
Mechanical Field Networks: Structured Neural Dynamics for Multivariate Systems
MF-Net learns a shared field state and mechanical transition rule from trajectories to deliver competitive forecasting and recoverable relation matrices on Lorenz-96 and real systems.
-
Beyond Isotropy in JEPAs: Hamiltonian Geometry and Symplectic Prediction
Fixed isotropic marginals in JEPAs can be maximally misaligned with unknown structured geometries, and HamJEPA using symplectic Hamiltonian leapfrog maps improves kNN and linear-probe performance on CIFAR-100 and Imag...
-
Symplectic Neural Operators for Learning Infinite Dimensional Hamiltonian Systems
Symplectic Neural Operators preserve symplectic structure for learning infinite-dimensional Hamiltonian PDEs and deliver improved long-term energy stability in theory and experiments.
-
PI-TTA: Physics-Informed Source-Free Test-Time Adaptation for Robust Human Activity Recognition on Mobile Devices
PI-TTA stabilizes source-free test-time adaptation for sensor-based human activity recognition by adding physics-consistent constraints, yielding up to 9.13% accuracy gains and lower physical violation rates on three ...
-
Area Scaling of Dynamical Degrees of Freedom in Regularised Scalar Field Theory
The minimal number of dynamical degrees of freedom in regularised scalar field theory scales with area, governed by the count of distinct normal-mode frequencies below the ultraviolet cutoff.
-
Gravitational Duals from Equations of State II: Large Hierarchies and False Vacua
Extends PINN-based holographic inverse problem solving to false vacuum and large hierarchy regimes, claiming accurate scalar potential reconstruction from thermodynamic data despite numerical stiffness.
-
Symplecticity-preserving prediction of parameter-dependent Hamiltonian dynamics by Generalized Kernel Interpolation
A product-kernel interpolation method is proposed that augments state with parameters to produce symplectic large-step predictors for Hamiltonian dynamics by construction, with error bounds that extend from the non-pa...
-
Robots Need More than VLA and World Models
The paper identifies four missing interfaces (data autolabelling, embodiment retargeting, physics-grounded world models, and video-based reward inference) as the central bottleneck beyond VLA scaling for robot intelligence.
-
Physically Viable World Models: A Case for Query-Conditioned Embodied AI
Embodied AI requires query-conditioned world models that select the simplest physical abstraction sufficient to answer intervention queries.
-
Physics-Informed Graph Neural Network Surrogates for Turbulent Nanoparticle Dispersion in Dental Clinical Environments
ELGIN is a graph-based physics-informed surrogate model that predicts carrier flow and polydisperse particle motion in dental aerosol scenarios, achieving lower tracking errors and 37x speedup versus full OpenFOAM CFD...
-
Physics-constrained identification of graph-based thermal networks for spacecraft digital twins
A physics-constrained inverse-problem framework identifies graph-based lumped-parameter thermal models from temperature measurements for spacecraft digital-twin applications.
-
Can Predicted Dynamics Exist in the Physical World?
Physical admissibility is defined as a prediction-control interface using kinematic, dynamic, and composed-horizon conditions to reject invalid dynamics proposals, with AUC 0.957 on LeRobot PushT and 87-89% prevention...
-
A comparative study of accuracy and rollout stability of temporal surrogate models
Comparative experiments on three chaotic systems find that architectures using integrator-like updates exhibit lower bias, reduced perturbation amplification, and more stable long-horizon rollouts than other common de...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.