Basics of Generative AI and LLMs CT-GENAI Practice Questions (Page 1)

Questions 1–5

1. 1expert · hard

   A test manager is considering using a generative AI model to create test data. The manager is concerned about the model producing data that is not realistic. What is the most effective way to address this concern?

   • ATrain the model on a large, representative dataset of real test data
   • BUse a rule-based system to generate the data instead
   • CUse a discriminative model to filter the generated data
   • DUse a larger model with more parameters
   Reveal Solution Discussion Select an answer first
2. 2application · medium

   A developer is using an LLM to autocomplete test code. The model suggests a function call that does not exist in the codebase. What is the most likely explanation?

   • AThe model is predicting the next token based on patterns in its training data, which may include code from other projects
   • BThe model has a bug in its tokenizer that corrupts the code context
   • CThe model is intentionally suggesting a new function that should be created
   • DThe model is overfitting to the current codebase and memorizing incorrect functions
   Reveal Solution Discussion Select an answer first
4. 3expert · hard

   A QA team is using an LLM to generate test cases. The team is concerned about the model producing test cases that are biased against certain user groups. What is the most effective way to mitigate this risk?

   • AReview the generated test cases for bias and adjust the prompts to be more inclusive
   • BUse a discriminative model to classify the test cases as biased or unbiased
   • CDisable the LLM's training data to prevent bias
   • DUse a larger LLM that is less likely to be biased
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   Which statement best describes a Large Language Model (LLM)?

   • AA neural network with billions of parameters trained on a diverse range of internet text.
   • BA database of all possible sentences in a language.
   • CA small model that runs on a single CPU and handles simple text classification.
   • DA rule-based system that uses a dictionary to translate words.
   Reveal Solution Discussion Select an answer first
6. 5application · easy

   A test team wants to use an LLM to translate API error messages from English into multiple languages for a global user base. Which LLM capability is being applied?

   • ATranslation
   • BSummarization
   • CText generation
   • DQuestion answering
   Reveal Solution Discussion Select an answer first