AI Driven Materials Discovery Tools - Universal Hyper-Active Learning for Machine Learning Interatomic Potentials - PSDI Webinar

The Physical Sciences Data Infrastructure (PSDI) aims to accelerate research in the physical sciences by providing a data infrastructure that brings together and builds upon the various data systems researchers currently use. Our webinar series provide updates on our exploratory pathfinder work, and on relevant tools and technologies that have been developed by members of our community.

PSDI is pleased to launch a new webinar series entitled “From Project to Platform: New Resources on PSDI”. This series aims to showcase the high-quality tools and resources developed through the funding call 2025, introduce them to a broader community, and foster engagement with relevant user groups. 

This record contains a more detailed version of the presentation that was presented at the session "AI Driven Materials Discovery Tools" - "Universal Hyper-Active Learning for Machine Learning Interatomic Potentials" on 30th April 2026. 

Professor James Kermode (University of Warwick) introduced the ase-uhal toolkit, which automates and accelerates the generation of high-quality training datasets for machine-learning interatomic potentials. The talk highlighted how a universal extension of the Hyperactive Learning framework improves data efficiency and model performance, enabling more accurate and scalable atomistic simulations.