Machine Learning / AI
In what form does LLM classify the query — what exactly does it return for scenario selection?
What is structured output and how is it used in LLM?
So, you've been looking for a job for 2 months now?
In terms of speed, which is faster — a regular agent with tools or a multi-agent system?
How do you describe tools? What is the general format and main requirements for description?
What is ReLU and why is it used?
Redis is non-persistent — what will happen if it restarts? Will we lose notifications?
Can you briefly tell what valuable things you learned there (in the project with appeals/OCR)?
Tell me about yourself: what have you been doing, what do you want to do, about your experience.
How were the coefficients for hybrid search selected?
Tell us what you know about search, ranking, recommendation, and personalization tasks.
Explain the differences between the dependency inversion principle and dependency injection in software development.
What is n in the context of attention complexity?
Explain why a two-tower network helps solve the first level ceiling problem — how exactly does it do that?
Don't you have some fixed salary part? Do you always receive the same salary?
Tell about gradient descent: what types are there?
Regarding location — do you consider remote work?
Tell me more about the project on employee churn prediction: what did you do, what data did you use, why did you choose CatBoost?
How do Gradient Clipping and Warmup help combat exploding gradients and loss divergence at the start of training?
Within the company (1C:Rarus), was there no opportunity to stay, grow, and develop in other areas?