Computational Notebooks like Jupyter or Google Collab interleave code with natural language documentation.

At the same time, Large language models allow us to automatically translate between code and natural language.

In this thesis, you will explore how we can combine these two aspects. The goal is to extend computational notebooks in such a way that code is no longer interleaved with additional documentation, which can lead to redudancies, but instead users can select for each cell whether they want to display a natural language summary of the code or its representation in a programming language of choice. You will then evaluate how this change affects developers.