Skip to navigation Skip to content
Careers | Phone Book | A - Z Index
Applied Computing 4 Scientific Discovery

Software, Libraries and Data

e3nn: A modular framework for Euclidean Neural Networks
Some of the most prevalent data types in the physical sciences and scientific computing are 3D geometry and geometric tensors. Euclidean neural networks have Euclidean symmetry built in and naturally handle geometry and geometric tensors as input and output.