An n-dimensional hybrid system embeds into a continuous vector field in m > 2n dimensions, enabling latent Neural ODEs with consistency losses to recover hybrid flows from time series.
hub
et al.Lagrangian Neural Networks
35 Pith papers cite this work, alongside 66 external citations. Polarity classification is still indexing.
35
Pith papers citing it
66
external citations · Pith
abstract
Accurate models of the world are built upon notions of its underlying symmetries. In physics, these symmetries correspond to conservation laws, such as for energy and momentum. Yet even though neural network models see increasing use in the physical sciences, they struggle to learn these symmetries. In this paper, we propose Lagrangian Neural Networks (LNNs), which can parameterize arbitrary Lagrangians using neural networks. In contrast to models that learn Hamiltonians, LNNs do not require canonical coordinates, and thus perform well in situations where canonical momenta are unknown or difficult to compute. Unlike previous approaches, our method does not restrict the functional form of learned energies and will produce energy-conserving models for a variety of tasks. We test our approach on a double pendulum and a relativistic particle, demonstrating energy conservation where a baseline approach incurs dissipation and modeling relativity without canonical coordinates where a Hamiltonian approach fails. Finally, we show how this model can be applied to graphs and continuous systems using a Lagrangian Graph Network, and demonstrate it on the 1D wave equation.
hub tools
citation-role summary
background 3
citation-polarity summary
roles
background 3representative citing papers
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 before the advantage closes near eleven steps.
GON uses 2-jet features and an anchor-and-variance objective to fix gauge freedom in ordinal predictability scoring, enabling pretrained initialization to outperform scratch training on held-out dynamical systems.
A Set-Transformer architecture with self-attention encodes Pauli-string correlations, optimizes via commutation objective, and finds symmetries with near-deterministic success on physical models like Ising and Toric code.
FISolver trains a compact LLM on backward-generated (differential equation, first integral) pairs and uses guided reinforcement learning to outperform larger models and Mathematica on first-integral benchmarks at lower cost.
Hamiltonian Transformer with norm-preserving attention and phase embeddings outperforms baselines in RF fingerprinting on WiSig dataset, reaching 99.12% same-day accuracy and 61.64% at 150 transmitters.
VHYDRO is a support-safe variational hybrid filter that jointly recovers continuous latent states, discrete contact modes, and sparse port-Hamiltonian laws per regime while preventing loss of feasible transitions.
HAAD detects deepfakes by modeling latent manifolds as potential energy surfaces and quantifying instability via Hamiltonian trajectory statistics such as action and energy dissipation.
piDMD learns a single parameter-affine Koopman surrogate ROM from training samples at multiple parameters to predict dynamics at unseen parameters with improved robustness over interpolation baselines.
SPADE adaptively enforces or relaxes structure priors in estimators via exact specification tests and Stein-unbiased James-Stein shrinkage with claimed oracle guarantees.
Neural ODEs constrained by the gradient of a jointly learned maximal Lyapunov function universally approximate locally exponentially stable dynamics within a region of attraction exactly given by the Lyapunov 1-sublevel set.
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.
LAPG combines conditional score-based diffusion with an action-derived guidance score to reduce phase drift and preserve physical invariants during temporal, parameter, and geometric extrapolation on free-fall, spring-mass, vortex, and airfoil 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.
NeuROK learns a data-driven latent kinematic parameterization on a large 4D dataset to generate realistic object deformations by simulating dynamics only in low-dimensional latent space via Lagrangian mechanics.
NHODE framework learns partially observed dynamical systems by combining Hamiltonian neural networks with neural ODEs, enforcing energy conservation and improving long-horizon stability over data-driven baselines on mass-spring and three-body problems.
ICDN is a neural network that models log-demand from log-prices so elasticities can be derived exactly by differentiation, showing better out-of-sample performance than log-log benchmarks on beer sales data.
Generative sequence models for physical tasks exhibit physical misgeneralization where local prediction errors propagate through physical measurements to distort aggregate distributions over quantities like distance or energy; a data deviation kernel explains and predicts the shifts and supports a内核
LaWM induces latent transitions from a learned discrete variational principle rather than an unconstrained neural predictor, yielding improved physical consistency on synthetic dynamics and robot benchmarks.
DiLaR-PINN learns dissipative effects in electromechanical systems via a skew-dissipative latent residual PINN that guarantees non-increasing energy and uses recurrent curriculum training for partial observations.
SLIDE is a deep learning estimator that truncates initial effects via complex eigenvalues of linearized equations to predict output sequences of damped multibody systems, reporting speedups up to several million times.
Geometric deep learning provides a unified mathematical framework based on grids, groups, graphs, geodesics, and gauges to explain and extend neural network architectures by incorporating physical regularities.
Lifting non-conservative, actuated, and contact-constrained robot dynamics into an exactly symplectic phase-space map yields state-of-the-art out-of-distribution autoregressive rollout error at low parameter and FLOP cost.
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.
citing papers explorer
-
Embedding Hybrid Systems into Continuous Latent Vector Fields
An n-dimensional hybrid system embeds into a continuous vector field in m > 2n dimensions, enabling latent Neural ODEs with consistency losses to recover hybrid flows from time series.
-
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 before the advantage closes near eleven steps.
-
Learning Transferable Predictability Representations
GON uses 2-jet features and an anchor-and-variance objective to fix gauge freedom in ordinal predictability scoring, enabling pretrained initialization to outperform scratch training on held-out dynamical systems.
-
Attention-based optimizer for symmetry finding
A Set-Transformer architecture with self-attention encodes Pauli-string correlations, optimizes via commutation objective, and finds symmetries with near-deterministic success on physical models like Ising and Toric code.
-
Learning First Integrals via Backward-Generated Data and Guided Reinforcement Learning
FISolver trains a compact LLM on backward-generated (differential equation, first integral) pairs and uses guided reinforcement learning to outperform larger models and Mathematica on first-integral benchmarks at lower cost.
-
Hamiltonian-Inspired Attention Mechanism for Scalable RF Transmitter Fingerprinting
Hamiltonian Transformer with norm-preserving attention and phase embeddings outperforms baselines in RF fingerprinting on WiSig dataset, reaching 99.12% same-day accuracy and 61.64% at 150 transmitters.
-
Support-Safe Variational Hybrid Filtering for Contact-Mode and Sparse-Law Recovery
VHYDRO is a support-safe variational hybrid filter that jointly recovers continuous latent states, discrete contact modes, and sparse port-Hamiltonian laws per regime while preventing loss of feasible transitions.
-
Detecting Deepfakes via Hamiltonian Dynamics
HAAD detects deepfakes by modeling latent manifolds as potential energy surfaces and quantifying instability via Hamiltonian trajectory statistics such as action and energy dissipation.
-
Parametric Interpolation of Dynamic Mode Decomposition for Predicting Nonlinear Systems
piDMD learns a single parameter-affine Koopman surrogate ROM from training samples at multiple parameters to predict dynamics at unseen parameters with improved robustness over interpolation baselines.
-
SPADE: Structure-Prior Adaptive Decision Estimation
SPADE adaptively enforces or relaxes structure priors in estimators via exact specification tests and Stein-unbiased James-Stein shrinkage with claimed oracle guarantees.
-
Locally Stable Neural ODEs with Characterized Region of Attraction
Neural ODEs constrained by the gradient of a jointly learned maximal Lyapunov function universally approximate locally exponentially stable dynamics within a region of attraction exactly given by the Lyapunov 1-sublevel set.
-
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.
-
Least-Action-Guided Diffusion for Physical Extrapolation
LAPG combines conditional score-based diffusion with an action-derived guidance score to reduce phase drift and preserve physical invariants during temporal, parameter, and geometric extrapolation on free-fall, spring-mass, vortex, and airfoil systems.
-
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.
-
NeuROK: Generative 4D Neural Object Kinematics
NeuROK learns a data-driven latent kinematic parameterization on a large 4D dataset to generate realistic object deformations by simulating dynamics only in low-dimensional latent space via Lagrangian mechanics.
-
Learning partially observed systems with neural Hamiltonian ordinary differential equations
NHODE framework learns partially observed dynamical systems by combining Hamiltonian neural networks with neural ODEs, enforcing energy conservation and improving long-horizon stability over data-driven baselines on mass-spring and three-body problems.
-
Integrable Elasticity via Neural Demand Potentials
ICDN is a neural network that models log-demand from log-prices so elasticities can be derived exactly by differentiation, showing better out-of-sample performance than log-log benchmarks on beer sales data.
-
Mechanisms of Misgeneralization in Physical Sequence Modeling
Generative sequence models for physical tasks exhibit physical misgeneralization where local prediction errors propagate through physical measurements to distort aggregate distributions over quantities like distance or energy; a data deviation kernel explains and predicts the shifts and supports a内核
-
LaWM: Least Action World Models for Long-Horizon Physical Consistency from Visual Observations
LaWM induces latent transitions from a learned discrete variational principle rather than an unconstrained neural predictor, yielding improved physical consistency on synthetic dynamics and robot benchmarks.
-
Dissipative Latent Residual Physics-Informed Neural Networks for Modeling and Identification of Electromechanical Systems
DiLaR-PINN learns dissipative effects in electromechanical systems via a skew-dissipative latent residual PINN that guarantees non-increasing energy and uses recurrent curriculum training for partial observations.
-
SLIDE: A machine-learning based method for forced dynamic response estimation of multibody systems
SLIDE is a deep learning estimator that truncates initial effects via complex eigenvalues of linearized equations to predict output sequences of damped multibody systems, reporting speedups up to several million times.
-
Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges
Geometric deep learning provides a unified mathematical framework based on grids, groups, graphs, geodesics, and gauges to explain and extend neural network architectures by incorporating physical regularities.
-
CaLiSym: Learning Symplectic Dynamics of Real-World Systems through Structured Canonical Lifts
Lifting non-conservative, actuated, and contact-constrained robot dynamics into an exactly symplectic phase-space map yields state-of-the-art out-of-distribution autoregressive rollout error at low parameter and FLOP cost.
-
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 in a preliminary single-case test.
-
Physically Native World Models: A Hamiltonian Perspective on Generative World Modeling
The paper introduces Hamiltonian World Models by encoding observations into structured latent phase space and evolving states via Hamiltonian-inspired dynamics for physically meaningful rollouts in embodied AI.
-
Mesh Field Theory: Port-Hamiltonian Formulation of Mesh-Based Physics
Mesh Field Theory proves mesh-based continuum physics reduces to port-Hamiltonian dynamics with topology fixing interconnections and metrics entering only via constitutive relations, enabling MeshFT-Net for stable, data-efficient simulation.
-
Optimal transport by a Lagrangian dynamics of population distribution
A quadratic Lagrangian incorporating dissipation models human mobility from population distributions and fits both synthetic and empirical data, showing comparable inertia and dissipation effects.
-
Geometry-Conditioned Fourier Neural Operators for Cubic Nonlinear Schrodinger Dynamics on Periodic Domains
A Fourier neural operator conditioned on the torus aspect-ratio parameter learns NLS dynamics and reproduces stronger H²-norm growth on rational tori than on irrational tori.
-
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 of invalid actions in interventions.
-
World Models: A Comprehensive Survey of Architectures, Methodologies, Reasoning Paradigms, and Applications
The paper delivers a multi-axis taxonomy for world models that maps architectures, training families, reasoning strategies, and domains from early cognitive foundations through systems such as Dreamer, MuZero, and Sora while noting evaluation gaps.
-
L-Learning : A Lyapunov-Based Approach Leveraging Lagrangian Mechanics for Efficient and Stable Robot Tracking
L-Learning learns the system's energy function from data to deliver stable, accurate, and sample-efficient robot trajectory tracking with closed-loop stability guarantees.
-
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 designs when model capacity is matched.
- Symplectic Neural Networks for Learning Non-Separable Hamiltonians