With this idea, I wanted to incorporate a kind of continuous value with using tokens such as lego blocks. Each of the individual blocks represents a row in the data with a specific dimension, and each block is scaled according to an average calculated from every single row. The problem with this idea is that it becomes impossible to see attributes of blocks inside, meaning that this model must have the ability to be split apart and viewed as slices/individual objects. However, it also makes comparing two objects easy. For example, if there is an object with a lot of blocks, but the scale of the object itself is relatively small in comparison, it demonstrates the idea that the average must not be high for that specific dimension, even with a lot of rows dedicated to it.