BiBERTa: a self-supervised framework for accelerating the discovery of stable organic photovoltaic materials
A novel bi-encoder architecture RoBERTa framework (BiBERTa) integrates self-supervised pretraining to accelerate the discovery of high-performance organic photovoltaic materials with limited experimental data (MAE = 1.67%).
Authors
- Yingping Zou (ORCID: https://orcid.org/0000-0003-1901-7243)
- Hongmei Lü (ORCID: https://orcid.org/0000-0002-4686-4491)
- Jie Zou (ORCID: https://orcid.org/0000-0001-6019-5194)
- Hailiang Zhang (ORCID: https://orcid.org/0000-0002-8050-2494)
- Dongxu Li (ORCID: https://orcid.org/0000-0002-4392-8672)
- Jinyu Sun
- Xiaoxiao Tan
- Yue Wang (ORCID: https://orcid.org/0009-0007-6163-5885)
- Zhimin Zhang (ORCID: https://orcid.org/0009-0009-3591-3890)
Institutions
- Central South University (CN)
Publication Details
- Journal
- Journal of Materials Chemistry A
- Published
- 2025-01-01
- DOI
- https://doi.org/10.1039/d5ta01529d
- Citations
- 1
- Primary Topic
- Machine Learning in Materials Science
- Type
- article
- Field-Weighted Citation Impact
- 0.52