Identify Hallucinations, Reasoning Errors and Biases in LLM Output CT-GENAI Practice Questions (Page 7)

Questions 31–35

1. 31expert · hard

   A test team uses an LLM to generate test scripts for a critical system. They have limited time and budget. They need to implement a mitigation strategy that addresses hallucinations, reasoning errors, and biases, but they cannot afford extensive manual review. Which strategy is most cost-effective?

   • AHire additional testers to manually review all generated scripts
   • BImplement automated validation tools that check facts, logic, and fairness
   • CLimit the LLM's output to simple test cases to avoid errors
   • DUse a more expensive, higher-quality LLM to generate the test scripts
   Reveal Solution Discussion Select an answer first
2. 32expert · hard

   A test team uses an LLM to generate test data for a healthcare application. The LLM produces a patient record with a diagnosis code 'Z99.9' that the team cannot find in the current ICD-10-CM manual. The team is divided: some believe it is a hallucination, others think it is a valid code from a newer edition. The team has limited time and must decide whether to use the code in test data. Which action is most appropriate?

   • AUse the code because the LLM is trained on recent medical data
   • BAsk the LLM to provide a source for the code
   • CVerify the code against the latest official ICD-10-CM coding manual
   • DReplace the code with a similar code from the current manual
   Reveal Solution Discussion Select an answer first
4. 33foundation · easy

   What is the purpose of a consistency check when evaluating LLM outputs?

   • ATo ensure the output is grammatically correct
   • BTo verify that the output is consistent with the prompt and does not contradict itself
   • CTo check if the output is formatted correctly
   • DTo measure the model's response time
   Reveal Solution Discussion Select an answer first
5. 34application · medium

   A QA team uses an LLM to generate test scenarios for a workflow engine. The LLM produces a complex test case with multiple conditional branches. The team wants to detect potential reasoning errors before executing the tests. Which approach is most effective?

   • ACompare the test case with a similar test case generated for another feature
   • BAsk the LLM to generate a simpler test case
   • CRun the test case once and check if it passes
   • DBreak the test case into individual steps and validate each step's logic against the requirements
   Reveal Solution Discussion Select an answer first
6. 35foundation · easy

   What is algorithmic amplification in the context of LLM biases?

   • AThe process by which the model's architecture intensifies biases present in the training data
   • BThe process of making the model faster
   • CThe process of making the model more accurate
   • DThe process of adding more data to reduce bias
   Reveal Solution Discussion Select an answer first