Large Language Models (LLMs) have revolutionized natural language processing by generating human-like text and images from textual input. However, their potential to generate complex 2D/3D visualizations has been largely unexplored. We report initial experiments showing that LLMs can generate 2D/3D visualizations that may be used for legal visualization. Further research is needed for complex 2D visualizations and 3D scenes. LLMs can become a powerful tool for many industries and applications, generating complex visualizations with minimal training.

翻译：大语言模型通过从文本输入生成类人文本与图像，已彻底革新了自然语言处理领域。然而，其生成复杂2D/3D可视化的潜力尚未得到充分探索。我们报告了初步实验，表明大语言模型能够生成可用于法律可视化的2D/3D可视化内容。针对复杂2D可视化与三维场景的研究仍需深入开展。大语言模型有望成为众多行业与应用领域的有力工具，仅需极少训练即可生成复杂可视化内容。