What is Double DQN and what problem does it solve?
Machine Learning / AI
What is calibration and why is it important for fairness?
What is dropout?
How do you limit the number of steps an agent takes so it doesn't get stuck?
What is interleaving in search and recommendations?
What is an inverted index and what data structures are used?
What is long-context evaluation? What is needle-in-haystack?
What happens to code from a Jupyter notebook afterwards?
What is shared memory and why use it?
What is stream-based active learning?
What is a skip connection and why is it needed (ResNet)?
Why is the normal distribution often used in statistics and machine learning?
What is contrastive learning on graphs (DGI, GraphCL)?
What features do voice dialogues have (latency, ASR errors)?
What is the asymptotic complexity of searching in a list with a nested loop?
What problems arise in multilingual RAG?
What are the features of active learning in computer vision?
What is hierarchical softmax and where is it used?
Design a support chatbot system based on RAG with a guarantee of no hallucination.
What is the difference between L1 and L2 regularization?