Mitigation of Non-Deterministic Behavior of LLMs CT-GENAI Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   What is the purpose of designing robust test assertions for LLM outputs?

   • ATo ensure the LLM always produces the exact same output
   • BTo make the test fail whenever the output changes
   • CTo increase the speed of test execution
   • DTo validate the correctness of the output while tolerating acceptable variations
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   Which API parameter is typically used to control the randomness of LLM outputs?

   • Amodel_version
   • Bmax_tokens
   • Ctemperature
   • Dstop_sequences
   Reveal Solution Discussion Select an answer first
4. 23foundation · easy

   What is a key benefit of controlling the deployment of LLM models in a test environment?

   • AIt allows testers to use the latest features immediately
   • BIt eliminates the need for monitoring LLM outputs
   • CIt prevents unexpected changes in model behavior from affecting test results
   • DIt reduces the need for test data preparation
   Reveal Solution Discussion Select an answer first
5. 24foundation · easy

   Which monitoring metric is most useful for detecting increased non-determinism in LLM outputs?

   • AThe variance in outputs for identical inputs across multiple runs
   • BThe response time of the API
   • CThe cost per API call
   • DThe number of tokens generated per request
   Reveal Solution Discussion Select an answer first
6. 25foundation · easy

   Which of the following is a primary source of non-deterministic behavior in LLM outputs?

   • AThe use of a fixed random seed for every generation
   • BThe model's training data being static and unchanging
   • CThe temperature parameter set to 0.0
   • DThe sampling method that selects tokens based on probability distributions
   Reveal Solution Discussion Select an answer first