Semantic tessellations are geometric patterns created by combining shapes to form a larger, repeating pattern. The shapes used to create the tessellations often have a symbolic meaning, making them visually interesting as well as thought-provoking. Semantic tessellations are commonly found in Islamic art, textiles, and architectural decorations, and can be used to create intricate and beautiful designs. In machine learning, semantic tessellations can be used to create a map of relationships between objects. By combining shapes that represent different entities, a visual image can be created that shows how those entities are related. This can provide insights into complex datasets and can be used to create models that can detect patterns and relationships in data. Hierarchical tesselations are a type of semantic tessellation which are created by combining shapes to form a tree-like structure. This structure can be used to represent the hierarchical relationship between objects, such as in an organizational chart or a family tree. Hierarchical tesselations can also be used to represent the structure of a network or other complex system. Hierarchical tesselations are a specific data structure composed of layers of embeddings, which are used to represent the relationships between different objects. The embeddings are created by combining shapes that have a specific meaning, allowing the hierarchical relationships between objects to be visualized. They can be used to create models that can detect patterns and relationships in data, allowing for more accurate decision making and analysis.