Utilization of Large Language Models for Materials Discovery
발표자
()
초록
내용
In this talk, I will talk about utilizing large language models (LLM) for materials search, with particular emphasis on metal-organic frameworks (MOFs). In particular, around 40,000 MOF papers were mined using LLM showing excellent accuracy compared to rule-based and machine-learning based text mining algorithms. The utility of using this data is explained, with specific emphasis on identifying the optimal experimental synthesis conditions for materials. Next, we introduce ChatMOF, which is an autonomous AI system for MOFs and display how inverse design is possible using ChatMOF and how LLM can be used to facilitate using these tools for novices.