Fine-tuning large language models to generate single-atom catalyst synthesis procedures
Manu Suvarna, Matteo Manica, Fillipo Ficarra, Andres M Bran, Andrea Ruiz-Ferrando, Magdalena Lederbauer, Philippe Schwaller, Javier Pérez-Ramírez, Teodoro Laino · 2026
Read Fine-tuning large language models to generate single-atom catalyst synthesis procedures by Manu Suvarna, Matteo Manica, Fillipo Ficarra, Andres M Bran, Andrea Ruiz-Ferrando, Magdalena Lederbauer, Philippe Schwaller, Javier Pérez-Ramírez, Teodoro Laino (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/PMC13639067.txt.
Description
Manu Suvarna and eight colleagues describe a 2026 project that fine-tunes a large language model to write recipes for single-atom catalysts. These catalysts, with isolated metal atoms on a support, are hard to design because knowledge sits scattered across fast-growing literature.
The team trained the model on 2,964 publications and built an interface where researchers request procedures by metal, support, and target reaction. Human reviewers checked 150 generated protocols and found the steps coherent and chemically sensible, though with clear limits. The authors stress that model output must be reviewed critically and used as an assistant, not an authority.
Catalog record
| CREATOR | Manu Suvarna, Matteo Manica, Fillipo Ficarra, Andres M Bran, Andrea Ruiz-Ferrando, Magdalena Lederbauer, Philippe Schwaller, Javier Pérez-Ramírez, Teodoro Laino |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 55381 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |