Machine Learning / AI

What is post-training quantization for mobile models?

What is model interpretability and what methods are there (SHAP, LIME, Integrated Gradients)?

What are Bayesian neural networks and how to compute the posterior?

What fairness issues does ML have in medicine?

Compare pre-norm and post-norm transformers. Which one trains more stably at greater depths?

What is a search budget and how to plan it?

What is NLU and what are its sub-tasks (intent, slots, entities)?