Advanced NLP Interview Questions, issue 4, Dec 11, 2025
The WEAT Bias Detection Trap
final round ML Engineer interview at Google DeepMind, and the interviewer asks:
“How do you prove your word embeddings aren’t biased before we ship?”
Why your embeddings look ‘mathematically perfect’ while secretly encoding 100 years of human bias.
The full answer, with the mechanism and the arithmetic, is for paid subscribers on Substack.