Module Quiz

Question 1

What is the core mechanism by which an LLM produces output?

Select one answer

It queries a structured knowledge base and returns matching facts

It predicts the most probable next token based on all preceding tokens

It executes a set of logical inference rules over a symbolic knowledge graph

It searches training data at inference time to find the closest matching example

Question 2

Which of the following best explains why an LLM may recommend a flag that does not exist in the mesheryctl CLI?

Select one answer

The model has access to an outdated version of the mesheryctl documentation

The model generates statistically plausible text regardless of whether the content is accurate

The model is deliberately testing whether the user will verify its output

The model misread the user's prompt and answered a different question

Question 3

A Meshery design with 10 components serializes to approximately 2,000 tokens. You also want to include a 3,000-token policy document and a 500-token system prompt. Which of the following statements is true?

Select one answer

Context window limits only apply to the model's output, not the input

The total input of 5,500 tokens leaves limited room for model output in smaller context windows

Token counts are irrelevant because modern models have unlimited context

Only the system prompt counts against the context window

Question 4

Which task types are well-suited to smaller, faster LLMs rather than large frontier models? Select all that apply.

Select all that apply

Classifying log lines as error, warning, or informational

Multi-step root cause analysis across a distributed system incident

Routing a support request to the correct team based on its content

Evaluating whether a proposed network policy satisfies multiple security constraints simultaneously