9 papers
Inverse design of bespoke interatomic potentials via active learning by information-matching
Yonatan Kurniawan, Logan D. Williams, Amit Samanta +6
Interatomic potentials (IPs) enable large-scale atomistic simulations beyond the reach of first-principles methods, but their predictive reliability depends critically on the selec…
Composable and adaptive design of machine learning interatomic potentials guided by Fisher-information analysis
Weishi Wang, Mark K. Transtrum, Vincenzo Lordi +2
An adaptive physics-inspired model design strategy for machine-learning interatomic potentials (MLIPs) is proposed. This strategy relies on iterative reconfigurations of composite…
An information-matching approach to optimal experimental design and active learning
Yonatan Kurniawan, Tracianne B. Neilsen, Benjamin L. Francis +7
The efficacy of mathematical models heavily depends on the quality of the training data, yet collecting sufficient data is often expensive and challenging. Many modeling applicatio…
An Analytical Characterization of Sloppiness in Neural Networks: Insights from Linear Models
Jialin Mao, Itay Griniasty, Yan Sun +3
Recent experiments have shown that training trajectories of multiple deep neural networks with different architectures, optimization algorithms, hyper-parameter settings, and regul…
A Computational Picture of Hydride Formation and Dissipation In Nb SRF Cavities
Aiden Harbick, Mark Transtrum, Nathan Sitaraman +2
Research linking surface hydrides to Q-disease, and the subsequent development of methods to eliminate surface hydrides, is one of the great successes of SRF cavity R&D. We use tim…
A Time-Dependent Ginzburg-Landau Framework for Sample-Specific Simulation of Superconductors for SRF Applications
Aiden V. Harbick, Mark K. Transtrum
Modern superconducting radio frequency (SRF) applications demand precise control over material properties across multiple length scales - from microscopic composition, to mesoscopi…