# TeaNet: universal neural network interatomic potential inspired by iterative electronic relaxations

arXiv: https://arxiv.org/abs/1912.01398 · HTML: https://mlsanity.com/paper/1912.01398/

## Key facts

- Cited by 16 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Dec 2019.

## Built on by (most cited first, 16 of 16)

- [Smooth, exact rotational symmetrization for deep learning on point clouds](https://mlsanity.com/paper/2305.19302/) (2023)
- [A practical guide to machine learning interatomic potentials -- Status and future](https://mlsanity.com/paper/2503.09814/) (2025)
- [Evaluation of the MACE Force Field Architecture: from Medicinal Chemistry to Materials Science](https://mlsanity.com/paper/2305.14247/) (2023)
- [CHGNet: Pretrained universal neural network potential for charge-informed atomistic modeling](https://mlsanity.com/paper/2302.14231/) (2023)
- [Tensor-reduced atomic density representations](https://mlsanity.com/paper/2210.01705/) (2022)
- [FeNNol: an Efficient and Flexible Library for Building Force-field-enhanced Neural Network Potentials](https://mlsanity.com/paper/2405.01491/) (2024)
- [Symmetry-adapted graph neural networks for constructing molecular dynamics force fields](https://mlsanity.com/paper/2101.02930/) (2021)
- [Crystal structure prediction using neural network potential and age-fitness Pareto genetic algorithm](https://mlsanity.com/paper/2309.06710/) (2023)
- [Dielectric Tensor Prediction for Inorganic Materials Using Latent Information from Preferred Potential](https://mlsanity.com/paper/2405.09052/) (2024)
- [Molecule Graph Networks with Many-body Equivariant Interactions](https://mlsanity.com/paper/2406.13265/) (2024)
- [Enhancing Machine Learning Potentials through Transfer Learning across Chemical Elements](https://mlsanity.com/paper/2502.13522/) (2025)
- [Multi-task learning for molecular electronic structure approaching coupled-cluster accuracy](https://mlsanity.com/paper/2405.12229/) (2024)
- [Neural Network Matrix Product Operator: A Multi-Dimensionally Integrable Machine Learning Potential](https://mlsanity.com/paper/2410.23858/) (2024)
- [Are Foundational Atomistic Models Reliable for Finite-Temperature Molecular Dynamics?](https://mlsanity.com/paper/2503.08207/) (2025)
- [LaMM: Semi-Supervised Pre-Training of Large-Scale Materials Models](https://mlsanity.com/paper/2505.22208/) (2025)
- [Benchmarking Universal Interatomic Potentials on Zeolite Structures](https://mlsanity.com/paper/2509.07417/) (2025)

## Abstract (arXiv)

A universal interatomic potential for an arbitrary set of chemical elements is urgently needed in computational materials science. Graph convolution neural network (GCN) has rich expressive power, but previously was mainly employed to transport scalars and vectors, not rank ≥ 2 tensors. As classic interatomic potentials were inspired by tight-binding electronic relaxation framework, we want to represent this iterative propagation of rank ≥ 2 tensor information by GCN. Here we propose an architecture called the tensor embedded atom network (TeaNet) where angular interaction is translated into graph convolution through the incorporation of Euclidean tensors, vectors and scalars. By applying the residual network (ResNet) architecture and training with recurrent GCN weights initialization, a much deeper (16 layers) GCN was constructed, whose flow is similar to an iterative electronic relaxation. Our traning dataset is generated by density functional theory calculation of mostly chemically and structurally randomized configurations. We demonstrate that arbitrary structures and reactions involving the first 18 elements on the periodic table (H to Ar) can be realized satisfactorily by TeaNet, including C-H molecular structures, metals, amorphous SiO2, and water, showing surprisingly good performance (energy mean absolute error 19 meV/atom) and robustness for arbitrary chemistries involving elements from H to Ar.