interatomic-potentials

NequIP is a code for building E(3)-equivariant interatomic potentials

🏆 A ranked list of awesome atomistic machine learning projects ⚛️🧬💎.

Allegro is a code for building highly scalable E(3)-equivariant interatomic potentials

[ICLR 2024] EquiformerV2: Improved Equivariant Transformer for Scaling to Higher-Degree Representations

[ICLR 2023 Spotlight] Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphs

SO3krates and Universal Pairwise Force Field for Molecular Simulation

Machine-Learned Interatomic Potential eXploration (mlipx) is designed at BASF for evaluating machine-learned interatomic potentials (MLIPs). It offers a growing set of evaluation methods alongside powerful visualization and comparison tools.

🌟 [NeurIPS '25 Spotlight] Fair and transparent benchmark of machine learning interatomic potentials (MLIPs), beyond basic error metrics https://openreview.net/forum?id=SAT0KPA5UO

python library for atomistic machine learning

A Python library and command line interface for automated free energy calculations

[ICML'26] Phonon fine-tuning (PFT) and [NeurIPS'25 AI4Mat] Nequix: Training a foundation model for materials on a budget

UF3: a python library for generating ultra-fast interatomic potentials

ML Performance and Extrapolation Guide (reference deployment https://ml-peg.stfc.ac.uk)

An open-source active learning framework for training machine-learned interatomic potentials

Allegro-pol extends the Allegro architecture to model the electric response of materials

KIM-based Learning-Integrated Fitting Framework for interatomic potentials.

A flexible and performant framework for training machine learning potentials.

FLAME: a library for atomistic modeling environments

PotentialLearning.jl: Optimize your atomistic data and interatomic potential models in your molecular dynamics workflows.

Generator of polynomial machine learning potentials