This article presents a new NLP task called structured information inference (SIS) to address the complexities of information extraction at the device level in materials science. We accomplished this task by finetuning GPT-3 on a exsiting perovskite solar cell FAIR dataset with 91.8 F1-score and we updated the dataset with all related scientific papers up to now. The produced dataset is formatted and normalized, enabling its direct utilization as input in subsequent data analysis. This feature will enable materials scientists to develop their own models by selecting high-quality review papers within their domain. Furthermore, we designed experiments to predict PCE and reverse-predict parameters and obtained comparable performance with DFT, which demonstrates the potential of large language models to judge materials and design new materials like a materials scientist.

翻译：本文提出一种名为结构化信息推理（SIS）的新型自然语言处理任务，旨在解决材料科学中器件级信息提取的复杂性。我们通过对现有钙钛矿太阳能电池FAIR数据集微调GPT-3模型实现了该任务，取得了91.8的F1分数，并将数据集更新至涵盖迄今所有相关科研论文。生成的数据集经过格式化和标准化处理，可直接用作后续数据分析的输入。这一特性将使材料科学家能够通过选择各自领域内高质量的综述论文来开发自己的模型。此外，我们设计了预测光电转换效率（PCE）及反向预测参数的实验，获得了与密度泛函理论（DFT）相当的性能，这表明大型语言模型具有像材料科学家一样评估材料并设计新材料的潜力。