Equivariant Transformers for Neural Network based Molecular Potentials.
Philipp ThölkeGianni De FabritiisPublished in: ICLR (2022)
Keyphrases
- neural network
- high order
- higher order
- three dimensional
- rotation invariant
- matrix valued
- real time
- molecular structures
- molecular structure
- data sets
- image processing
- case study
- dna computing
- potential functions
- partial discharge
- drug design
- probabilistic model
- pairwise
- multiscale
- social networks
- learning algorithm
- machine learning