Benchmarking Compositional Generalisation for Machine Learning Interatomic Potentials
NeurIPS 2026
Benchmarking Compositional Generalisation for Machine Learning Interatomic Potentials
NeurIPS 2026
For chemists: Knowing the parts, missing the whole
Seminar version. Leads with the chemistry (functional groups, molecular dynamics, DFT) and keeps machine learning jargon light.
For computer scientists: Knowing the parts, missing the whole
Seminar version. Leads with compositional generalisation and out-of-distribution evaluation, and explains the chemistry from scratch.
Explainable Graph Neural Network for Computational Chemistry
First-year review · Cardiff University · 2025