100% Free ISTQB CT-GenAI Practice Test Questions

Question 1

Which option BEST differentiates the three prompting techniques?

A : Few-shot = no examples; Chaining = single prompt; Meta = disable iteration

B : Meta = step decomposition; Chaining = zero-shot only; Few-shot = manual optimization

C : Chaining = give examples; Few-shot = break tasks; Meta = manual edits only

D : Few-shot = examples; Chaining = multi-step prompts; Meta = model helps draft/refine prompts

Answer: D

Question 2

4.2.2 An organization develops an in-house LLM-based test automation framework. While privacy and control are maximized, the QA lead finds inconsistencies in how outputs are validated and maintained over time. Which LLMOps practice is essential to address this?

A : Establishing structured processes for validation and maintenance.

B : Outsourcing validation tasks to the LLM-as-a-Service provider.

C : Reducing dataset size to lower validation effort.

D : Limiting model retraining cycles to once per year.

Answer: A

Question 3

4.2.1 During fine-tuning of an LLM for test generation, the training dataset contains inconsistencies: some user stories are incomplete, and some test cases do not match the described functionality. After fine-tuning, the model frequently generates irrelevant or biased test cases. Which fine-tuning challenge does this illustrate?

A : Overfitting caused by small dataset size.

B : Opacity in the model’s reasoning.

C : Computational overhead in processing large models.

D : Poor training data quality.

Answer: D

Question 4

4.2.2 An organization develops an in-house LLM-based test automation framework. While privacy and control are maximized, the QA lead finds inconsistencies in how outputs are validated and maintained over time. Which LLMOps practice is essential to address this?

A : Establishing structured processes for validation and maintenance.

B : Outsourcing validation tasks to the LLM-as-a-Service provider.

C : Reducing dataset size to lower validation effort.

D : Limiting model retraining cycles to once per year.

Answer: A

Question 5

1.1.2 A performance test uses an SLM to provide quick insights on defect clustering. The model’s outputs are fast but miss subtle context in multi-sentence defect reports. The team wants minimal increase in processing time while improving comprehension. Which change is most suitable?

A : Replace SLM with LLM despite increased computational time

B : Integrate embeddings fine-tuned on historical defect data

C : Expand tokenization to the smallest possible tokens

D : Increase the model’s temperature parameter for output variety

Answer: B

100% Free CT-GenAI Practice Exam Questions and Answers