This set contains seven independent multiple-choice questions from the ML 2.0 section of the assessment. Each question is self-contained: read the stem and pick the single best option.

Topics covered: retrieval-augmented generation (RAG) and the session/context model of an MCP server.

Question 1. How does poor Information Retrieval affect RAG model performance?

It leads to irrelevant or inaccurate answers
It decreases model size
It increases model training time
It doesn't affect the model

Question 2. What happens when context is lost in an MCP server session?

The session is terminated
The user receives a timeout message
The session continues without context
The user is notified to re-authenticate

Question 3. What is a primary challenge when integrating knowledge into RAG models?

Ensuring retrieved data is both relevant and accurate
Increasing model speed
Decreasing the number of data sources
Ensuring overfitting does not occur

Question 4. What is a critical step when fine-tuning a generative model for RAG?

Using a task-specific dataset
Using data from random sources
Ignoring model parameters
Using smaller datasets

Question 5. What is a key challenge in Information Retrieval for RAG?

Speed of response
Ensuring retrieved data is relevant and accurate
Reducing model size
Increasing model complexity

Question 6. Why is performance evaluation crucial in RAG models?

To ensure that the model generates relevant, accurate responses
To evaluate retrieval speed
To check the complexity of the model
To measure data size

Question 7. What is the cause of context drift in MCP server sessions?

Failure to recognize new input
Server overload
Loss of data from the session store
Inconsistent user behavior

Input Format

There is no input. Standard input is empty.

Output Format

Print seven lines. Line i must contain the option number (an integer from 1 to 4) that you have chosen for Question i.

Sample output shape

(This is only the shape of the expected output, not the answer key.)