96‐3: <i>Distinguished Student Paper:</i> Developing Large Language Models for Display Industrial Knowledge: Data Augmentation, Training Techniques, and Evaluation Strategies

Large Language Models (LLMs) hold numerous potential applications within the display industry. However, mainstream LLMs generally lack the domain‐specific knowledge. Especially, in display knowledge question‐answering(Q&amp;A) scenarios, the lack of understanding of specialized terminology leads to low accuracy in responses. Consequently, this paper proposes a development framework for a Display Industry Knowledge Large Language Model (DIK‐LLM), aimed at enhancing the model's comprehension of semiconductor display industry field through specialized data governance, knowledge distillation, data augmentation strategies, and continual pre‐training mechanisms. This approach not only significantly improves the model's performance in Q&amp;A applications within the display industry but also prevents catastrophic forgetting. It is hoped that these contents will provide guidance and reference for researchers and practitioners in the customization of LLM for specialized domains.