Kinetic predictions for S <sub>N</sub> 2 reactions using the BERT architecture: comparison and interpretation
This study introduces a BERT model for predicting experimental log k values of S N 2 reactions, comparing it to a Random Forest model. Both models achieve similar accuracy and identify key chemical features.
Authors
- Stamatia Zavitsanou (ORCID: https://orcid.org/0000-0002-0571-2264)
- Fernanda Duarte (ORCID: https://orcid.org/0000-0002-6062-8209)
- Jason Crain (ORCID: https://orcid.org/0000-0001-8672-9158)
- James David Somper
- Maria Calvo
- Ewa Wieczorek
- Chloe Wilson
- Tom Watts
Institutions
- Mansfield University (US)
- Daresbury Laboratory (GB)
- University of Oxford (GB)
- Sci-Tech Daresbury (GB)
Publication Details
- Journal
- Digital Discovery
- Published
- 2025-12-26
- DOI
- https://doi.org/10.1039/d5dd00192g
- Primary Topic
- Advanced Chemical Physics Studies
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- University of Oxford
- Engineering and Physical Sciences Research Council