Machine Learning / AI
What is back-translation and how is it used in NMT training?
What are counterfactual explanations?
How do you deal with monotonous tasks?
What is intermittent demand forecasting (Croston)?
What data from call logs are stored: metadata or audio/video recordings?
Какая сложность по памяти и по времени у self-attention?
In which case should you choose Data Vault instead of third normal form for the core warehouse?
How do you verify that data from the source has reached the storage?
What are Bayesian neural networks and how to compute the posterior?
What is an IP-Adapter?
What is data retention?
How to merge data from two databases into a single model, such as a snowflake or star schema?
What protections are used (input/output filters, system prompt hardening, guardrails)?
When is cross-encoder used, and when is bi-encoder used in search?
What are you looking for in a job?