Evaluation Judges and Ground Truth GENERATIVE-AI-ENGINEER-ASSOCIATE Practice Questions (Page 3)

Questions 11–15

1. 11foundation · easy

   Which evaluation judge can be used when no ground truth labels are available?

   • ASelf-evaluation (LLM critiques its own output)
   • BROUGE score
   • CExact match
   • DBLEU score
   Reveal Solution Discussion Select an answer first
2. 12foundation · easy

   Which of the following evaluation methods depends on ground truth reference answers?

   • AModel confidence score
   • BROUGE score
   • CLLM-as-judge with no reference answer
   • DSelf-consistency check (model repeats its answer)
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   When selecting an evaluation judge for a task, why is the presence of ground truth important?

   • AIt determines the training data size
   • BIt determines whether you can use reference-based metrics or must use reference-free metrics
   • CIt determines the model architecture you should use
   • DIt determines the inference latency of the model
   Reveal Solution Discussion Select an answer first
5. 14application · medium

   A team is evaluating a text summarization model. They have a set of source articles with human-written reference summaries. They want to measure the quality of the generated summaries. Which evaluation judge would be most appropriate?

   • AROUGE
   • BReference-free LLM-as-judge
   • CExact match
   • DSelf-evaluation by the model
   Reveal Solution Discussion Select an answer first
6. 15foundation · easy

   Which of the following is an example of an evaluation judge that does NOT require ground truth?

   • AReference-free LLM-as-judge (evaluates output quality without a reference answer)
   • BBLEU
   • CROUGE
   • DExact match
   Reveal Solution Discussion Select an answer first