Asks a couple of behaviour questions based on the Leadership principles. Then basics of ML: ensemble methods, unsupervised methods, ROC-AUC, covering the breadth. Then for the depth: workings of RNN, Transformers, LLMs. A general coding problem, probably a Leetcode easy.
Two Sum question.
Why does random forest work so well?
Tell me about a time you disagreed with your boss/mentor and how you handled it
How to train the first mono-phone acoustic model?
Programming question: find the number of connected components from 2-dimensional grid (mark 1s as component, 0s for others).
Open question: can you describe some challenging problems you encounter and how do you solve it later?
Most familiar ML models
How to find minimum value in an array
Which loss function you'll use for a N-way classification problem?
Viewing 11 - 20 interview questions