AI finally figured out the messy, chaotic way atoms are packed inside glass, solving a mystery that's stumped scientists for decades.
March 25, 2026
Original Paper
Generative Inversion of Spectroscopic Data for Amorphous Structure Elucidation
arXiv · 2603.23210
The Takeaway
While we can easily map the neat rows of atoms in a crystal, the jumbled atoms in substances like glass or amorphous ice have been nearly impossible to visualize. This generative framework acts like a 'decoder,' turning messy experimental data into clear 3D maps of how these chaotic substances are actually built.
From the abstract
Determining atomistic structures from characterization data is one of the most common yet intricate problems in materials science. Particularly in amorphous materials, proposing structures that balance realism and agreement with experiments requires expert guidance, good interatomic potentials, or both. Here, we introduce GLASS, a generative framework that inverts multi-modal spectroscopic measurements into realistic atomistic structures without knowledge of the potential energy surface. A score