Abstract: The ability to apply advanced machine learning (ML) to process large amounts of heterogenous data can greatly aid in the understanding, discovery, and design of new materials for advanced application like clean energy, sustainable semiconductor manufacturing and drug discovery. The heterogenous nature of data encompasses materials structural data, material property data and qualitative descriptions across various types of materials, including solid-state (e.g., semiconductors, batteries), small molecules and proteins. This multitude of multi-modal data types necessitates the application of a diverse of advanced ML techniques across different interdisciplinary fields. Applying ML to materials discovery in a holistic manner can help accelerate material discovery in an end-to-end fashion, including faster hypothesis formulation with advanced language models, speeding up large-scale materials simulation by 10-100x and providing greater automation and sample efficiency for real-world experiments.
In this talk, I will present an overview of Intel Labs’ research efforts focused on applying ML to materials discovery, including materials property modeling using geometric deep learning, training scientific language models for materials science and applying generative AI to materials discovery.
Bio: Santiago Miret is a research lead at Intel Labs, focusing on applying AI to scientific problems with an emphasis on materials discovery and materials understanding. Santiago manages a wide range of academic collaborations focused on applying AI for scientific applications, including notable engagements with the Matter Lab led by Alán Aspuru-Guzik at the University of Toronto, and AI luminary Yoshua Bengio through multiple active collaborations at Mila in Montreal, both of which have led to cross-institutional publications at various machine learning venues.
Santiago is also a technical lead for AWASES, a new European research center on applying AI for sustainable semiconductor manufacturing that is jointly run by Intel and Merck KgaA, Darmstadt, Germany. In addition to his technical work, Santiago is leading efforts to build a strong research community on AI for materials design, including the first and second at NeurIPS 2022 and NeurIPS 2023. AI4Mat brings together domain experts from various fields of materials science and AI to exchange research work and ideas in an interdisciplinary forum.