Saved in:
Bibliographic Details
Main Authors: Deng, Congyue, Feng, Brandon Y., Garraffo, Cecilia, Garbarz, Alan, Walters, Robin, Freeman, William T., Guibas, Leonidas, He, Kaiming
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.07596
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912268350914560
author Deng, Congyue
Feng, Brandon Y.
Garraffo, Cecilia
Garbarz, Alan
Walters, Robin
Freeman, William T.
Guibas, Leonidas
He, Kaiming
author_facet Deng, Congyue
Feng, Brandon Y.
Garraffo, Cecilia
Garbarz, Alan
Walters, Robin
Freeman, William T.
Guibas, Leonidas
He, Kaiming
contents Machine learning frameworks for physical problems must capture and enforce physical constraints that preserve the structure of dynamical systems. Many existing approaches achieve this by integrating physical operators into neural networks. While these methods offer theoretical guarantees, they face two key limitations: (i) they primarily model local relations between adjacent time steps, overlooking longer-range or higher-level physical interactions, and (ii) they focus on forward simulation while neglecting broader physical reasoning tasks. We propose the Denoising Hamiltonian Network (DHN), a novel framework that generalizes Hamiltonian mechanics operators into more flexible neural operators. DHN captures non-local temporal relationships and mitigates numerical integration errors through a denoising mechanism. DHN also supports multi-system modeling with a global conditioning mechanism. We demonstrate its effectiveness and flexibility across three diverse physical reasoning tasks with distinct inputs and outputs.
format Preprint
id arxiv_https___arxiv_org_abs_2503_07596
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Denoising Hamiltonian Network for Physical Reasoning
Deng, Congyue
Feng, Brandon Y.
Garraffo, Cecilia
Garbarz, Alan
Walters, Robin
Freeman, William T.
Guibas, Leonidas
He, Kaiming
Machine Learning
Artificial Intelligence
Machine learning frameworks for physical problems must capture and enforce physical constraints that preserve the structure of dynamical systems. Many existing approaches achieve this by integrating physical operators into neural networks. While these methods offer theoretical guarantees, they face two key limitations: (i) they primarily model local relations between adjacent time steps, overlooking longer-range or higher-level physical interactions, and (ii) they focus on forward simulation while neglecting broader physical reasoning tasks. We propose the Denoising Hamiltonian Network (DHN), a novel framework that generalizes Hamiltonian mechanics operators into more flexible neural operators. DHN captures non-local temporal relationships and mitigates numerical integration errors through a denoising mechanism. DHN also supports multi-system modeling with a global conditioning mechanism. We demonstrate its effectiveness and flexibility across three diverse physical reasoning tasks with distinct inputs and outputs.
title Denoising Hamiltonian Network for Physical Reasoning
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2503.07596