papers

Publications (13)

cond-mat.dis-nn2022

Quantifying Disorder One Atom at a Time Using an Interpretable Graph Neural Network Paradigm

James Chapman, Tim Hsu, Xiao Chen +2

Quantifying the level of atomic disorder within materials is critical to understanding how evolving local structural environments dictate performance and durability. Here, we lever…

cond-mat.mtrl-sci2023

Spectroscopy-Guided Discovery of Three-Dimensional Structures of Disordered Materials with Diffusion Models

Hyuna Kwon, Tim Hsu, Wenyu Sun +8

The ability to rapidly develop materials with desired properties has a transformative impact on a broad range of emerging technologies. In this work, we introduce a new framework b…

cs.LG2022

Efficient, Interpretable Graph Neural Network Representation for Angle-dependent Properties and its Application to Optical Spectroscopy

Tim Hsu, Tuan Anh Pham, Nathan Keilbart +6

Graph neural networks are attractive for learning properties of atomic structures thanks to their intuitive graph encoding of atoms and bonds. However, conventional encoding does n…

math.GR2008

Artin HNN-extensions virtually embed in Artin groups

Tim Hsu, Ian J. Leary

An Artin HNN-extension is an HNN-extension of an Artin group in which the stable letter conjugates a pair of suitably chosen subsets of the standard generating set. We show that so…

physics.comp-ph2024

Score dynamics: scaling molecular dynamics with picoseconds timestep via conditional diffusion model

Tim Hsu, Babak Sadigh, Vasily Bulatov +1

We propose score dynamics (SD), a general framework for learning accelerated evolution operators with large timesteps from molecular-dynamics simulations. SD is centered around sco…

cs.LG2025

BOOM: Benchmarking Out-Of-distribution Molecular Property Predictions of Machine Learning Models

Evan R. Antoniuk, Shehtab Zaman, Tal Ben-Nun +9

Data-driven molecular discovery leverages artificial intelligence/machine learning (AI/ML) and generative modeling to filter and design novel molecules. Discovering novel molecules…