figshare
Browse

GPT synthesized narratives for materials science

Version 3 2024-06-04, 21:30
Version 2 2024-04-24, 17:51
Version 1 2024-03-10, 15:55
dataset
posted on 2024-06-04, 21:30 authored by YANG JEONG PARKYANG JEONG PARK

We provide the dataset including GPT-synthesized narratives based on the publicly accessible databases, Materials Project, JARVIS, OQMD, and AFLOW2. The dataset is designed to train future artificial intelligence (AI) models which aware the materials science.

Texts describe each material based on their properties to minimize the hallucination of chatbots. Furthermore, leveraging the chatbots' reasoning power, the texts also describe potential applications of each material. We believe that our dataset plays an important role in training future AI models.

History

Usage metrics

    Licence

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC