Interview Guides
You are reviewing two common ways to turn text into numeric features for an NLP model. One approach uses TF-IDF, the other uses word embeddings. Your goal is to explain how they represent text and why they behave differently in downstream tasks.
What are word embeddings, and how do they differ from TF-IDF?