General-purpose LLMs as Constrained Crystal Composition Generators
Hedda Oschinski, Maximilian L. Ach, Konstantin S. Jakob, Christian Carbogno, Karsten Reuter · 2026
Read General-purpose LLMs as Constrained Crystal Composition Generators by Hedda Oschinski, Maximilian L. Ach, Konstantin S. Jakob, Christian Carbogno, Karsten Reuter (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-31495.txt.
Description
Oschinski and colleagues test general-purpose language models as makers of constrained crystal compositions, using Elpasolite materials as a benchmark. Finding new inorganic materials is hard because the space of possible mixes is vast and full screening is costly, while task-specific generators need tedious training data first. With an iterative prompt-and-response loop, the models recover on average 96 percent of low-energy Elpasolites in the target region, beating prior task-specific models.
Catalog record
| CREATOR | Hedda Oschinski, Maximilian L. Ach, Konstantin S. Jakob, Christian Carbogno, Karsten Reuter |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 39614 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.31495