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：大型语言模型（LLMs）通过从文本输入生成类人文本和图像，彻底改变了自然语言处理领域。然而，它们生成复杂2D/3D可视化的潜力很大程度上尚未被探索。我们报告了初步实验，表明LLMs能够生成可用于法律可视化的2D/3D可视化。对于复杂的2D可视化和3D场景，仍需进一步研究。LLMs有望成为许多行业和应用中的强大工具，在极少训练下即可生成复杂可视化。