# Differentiable Physics-informed Graph Networks

arXiv: https://arxiv.org/abs/1902.02950 · HTML: https://mlsanity.com/paper/1902.02950/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 8 Feb 2019.

## Built on by (most cited first, 25 of 25)

- [Interpretable Machine Learning: Fundamental Principles and 10 Grand Challenges](https://mlsanity.com/paper/2103.11251/) (2021)
- [Multimodal learning with graphs](https://mlsanity.com/paper/2209.03299/) (2022)
- [Learning to schedule job-shop problems: Representation and policy learning using graph neural network and reinforcement learning](https://mlsanity.com/paper/2106.01086/) (2021)
- [Latent Space Data Assimilation by using Deep Learning](https://mlsanity.com/paper/2104.00430/) (2021)
- [Physics-Informed Machine Learning for Modeling and Control of Dynamical Systems](https://mlsanity.com/paper/2306.13867/) (2023)
- [Scalable algorithms for physics-informed neural and graph networks](https://mlsanity.com/paper/2205.08332/) (2022)
- [Climate Modeling with Neural Diffusion Equations](https://mlsanity.com/paper/2111.06011/) (2021)
- [PhyMPGN: Physics-encoded Message Passing Graph Network for spatiotemporal PDE systems](https://mlsanity.com/paper/2410.01337/) (2024)
- [Graph Neural Network-based surrogate model for granular flows](https://mlsanity.com/paper/2305.05218/) (2023)
- [Neural Pharmacodynamic State Space Modeling](https://mlsanity.com/paper/2102.11218/) (2021)
- [Physics-aware Machine Learning Revolutionizes Scientific Paradigm for Machine Learning and Process-based Hydrology](https://mlsanity.com/paper/2310.05227/) (2023)
- [PhyGNNet: Solving spatiotemporal PDEs with Physics-informed Graph Neural Network](https://mlsanity.com/paper/2208.04319/) (2022)
- [Neural Weighted A*: Learning Graph Costs and Heuristics with Differentiable Anytime A*](https://mlsanity.com/paper/2105.01480/) (2021)
- [Few-Sample Traffic Prediction with Graph Networks using Locale as Relational Inductive Biases](https://mlsanity.com/paper/2203.03965/) (2022)
- [Beyond Grid Data: Exploring Graph Neural Networks for Earth Observation](https://mlsanity.com/paper/2411.03223/) (2024)
- [Model discovery in the sparse sampling regime](https://mlsanity.com/paper/2105.00400/) (2021)
- [Physics-Coupled Spatio-Temporal Active Learning for Dynamical Systems](https://mlsanity.com/paper/2108.05385/) (2021)
- [Physics-Informed Graph Learning](https://mlsanity.com/paper/2202.10679/) (2022)
- [GN-SINDy: Greedy Sampling Neural Network in Sparse Identification of Nonlinear Partial Differential Equations](https://mlsanity.com/paper/2405.08613/) (2024)
- [ST-PCNN: Spatio-Temporal Physics-Coupled Neural Networks for Dynamics Forecasting](https://mlsanity.com/paper/2108.05940/) (2021)
- [AI Research Associate for Early-Stage Scientific Discovery](https://mlsanity.com/paper/2202.03199/) (2022)
- [Physics-Guided Adversarial Machine Learning for Aircraft Systems Simulation](https://mlsanity.com/paper/2209.03431/) (2022)
- [Geometric Autoencoder Priors for Bayesian Inversion: Learn First Observe Later](https://mlsanity.com/paper/2509.19929/) (2025)
- [Unraveled Multilevel Transformation Networks for Predicting Sparsely-Observed Spatiotemporal Dynamics](https://mlsanity.com/paper/2203.08655/) (2022)
- [Global Prediction of COVID-19 Variant Emergence Using Dynamics-Informed Graph Neural Networks](https://mlsanity.com/paper/2401.03390/) (2024)

## Abstract (arXiv)

While physics conveys knowledge of nature built from an interplay between observations and theory, it has been considered less importantly in deep neural networks. Especially, there are few works leveraging physics behaviors when the knowledge is given less explicitly. In this work, we propose a novel architecture called Differentiable Physics-informed Graph Networks (DPGN) to incorporate implicit physics knowledge which is given from domain experts by informing it in latent space. Using the concept of DPGN, we demonstrate that climate prediction tasks are significantly improved. Besides the experiment results, we validate the effectiveness of the proposed module and provide further applications of DPGN, such as inductive learning and multistep predictions.