Felix Morgers idéseminarium

With the advent of deep learning, models in natural language processing have become increasingly large and complex. As such, interpreting the decisions and internal workings of these models has become a more difficult yet important task. In this seminar, I will give an overview of how interpretability research has been done in natural language processing and elsewhere, as well as present the different research directions I am pursuing to better understand the internal workings of deep natural language processing systems.