2D Tokenization for Large Language Models

This article is about how we process text before passing it to large language models, the problems with this approach, and an alternative solution.