Machine Learning / AI

What is an MTP block in LLM?

Tell us about yourself.

How did you measure success in the [company] project (code autocompletion)? What metrics did you use?

What was the most difficult week at [company]?

What is tokenization?

Do you have experience designing and implementing complex neural network architectures (transformers, RNN)?

Name several types of classification metrics and explain them.

When do you choose full batch gradient descent vs mini-batch vs stochastic gradient descent? Why are there different options?

Is your experience more related to designing models from scratch or fine-tuning pre-trained models?

How do you reduce hallucinations in the information extraction pipeline?

What did you do at [company]?

What is a streaming API?

What does the company [company] do and what is your role there?

How can LLM be used for classification tasks and how to obtain class probabilities using LLM?

Describe a project at [company]: business problem, your role, how success was measured.

How does an LLM actually decide which tool to invoke?

Is Python your main language? How many years of commercial experience do you have?

How to increase inference speed for OpenRouter models through an AI-agent pipeline?

Tell us about your career overview: roles, industries, key projects.

Explain the terms 'training set' and 'test set' in the context of machine learning.