Research resources

Software & Data

Our research develops methods and tools for learning atomic interactions and studying reactive materials. The resources below connect model architectures, equivariant operations and atomistic simulation. Follow the linked repositories and papers for the current scope and availability of each project.

TACE Model Family

A family of equivariant atomistic models built around tensor representations and flexible physical embeddings, including TACE, TECE and magnetic TACE (mTACE).

TACE: Equivariant atomistic modeling with Cartesian tensor representations and flexible physical embeddings.
TECE: An edge cluster expansion with radial rotary attention for atomistic modeling.
mTACE: Magnetic atomistic models using local O(2) operators, with separate treatments for systems with and without spin–orbit coupling.

TACE Documentation TACE Preprint TECE Preprint mTACE Documentation mTACE and Local O(2) Preprint Source Code

TACE Foundations

Pretrained TACE-family models for atomistic simulation and fine-tuning. The model hub and usage guide describe the available releases and their intended applications.

Model Hub Usage Guide

EquivariantX

A library of equivariant operations for atomistic modeling, bringing together Cartesian tensor tools and local O(2) representation maps, frame transformations and operators.

cartnn / Cartesian-3j: Reference implementations of irreducible Cartesian tensor operations and equivariant interatomic architectures accompanying the Cartesian-3j study.
Local O(2) Operators: Composable representation maps and frame transformations that account for rotations and reflections, with time-reversal labels for magnetic quantities.

The cartnn code is a research reference implementation; consult its repository guidance before practical use.

cartnn Source Code Cartesian-3j ICML Paper Local O(2) Framework Preprint

REICO

An interaction-based approach to generating training data for general reactive element-based potentials. The linked studies address transferability and model construction for heterogeneous catalysis.

Nature Catalysis Paper Training Guide Data-Centric Perspective

Research Data

Data associated with individual studies should be accessed through the availability statements and repositories linked from the corresponding publications. For questions about a particular dataset or model, please contact Wenbo Xie.

wenboxie@cityu.edu.hk