Free NVIDIA NCA-GENL Actual Exam Questions - Question 11 Discussion

Question No. 11
You have developed a deep learning model for a recommendation system. You want to evaluate the
performance of the model using A/B testing. What is the rationale for using A/B testing with deep
learning model performance?
Select one option, then reveal solution.
US
IW
Irfan W.
2026-02-19

A imo, since A/B testing is about comparing real user impact, not model robustness or latency.

0
AY
Ahmed Y.
2026-02-10

A, because it’s about comparing user outcomes between two model versions.

0
MF
Mohammad F.
2026-02-02

It’s A because A/B testing focuses on comparing user responses to different model versions, so you can see which one actually improves recommendations in practice, not just in theory.

0
OA
Omar A.
2026-01-31

A/D? A seems solid since A/B testing is about direct comparison, but D also makes sense because latency can impact user experience, and measuring it could be part of evaluating model performance. Even if not the main goal, collecting latency data during A/B testing is practical. Options B and C feel off since B talks about rationale from designers—which isn’t really what A/B testing is for—and C talks about robustness, which is more about training and validation than A/B testing itself.

0
OA
Omar A.
2026-01-30

Makes sense to go with A since it directly compares two models’ user impact. A

0
SP
Sohail P.
2026-01-23

I think B and C can be ruled out since A/B testing is more about live user impact rather than internal model details or input variation. D mentions latency, but that’s just one aspect, not the core reason for A/B testing here.

0
SP
Sohail P.
2026-01-22

It’s A because A/B testing’s main goal is to compare user responses to different models in a live environment, which really shows which one works best in practice, not just in theory. D focuses too much on latency, which is just one part of the picture.

0
ZD
Zain D.
2026-01-19

D imo, latency comparison is a key part of evaluating model performance in real scenarios. While A is true, measuring response time directly impacts user experience, which A/B testing captures well.

0
AX
Ahmed X.
2026-01-16

A. Besides just comparing which model performs better, A/B testing also simulates real user interaction, which is crucial for recommendation systems. Options B, C, and D feel off because they focus on technical explanations or latency, which are not the main point of A/B testing here. It's really about measuring actual user response to different versions.

0
AX
Ahmed X.
2026-01-14

I think the answer is A. A/B testing is mainly about comparing two versions directly to see which one performs better in a real setting. The rest don't seem to fit the core purpose of A/B testing. Anyone else agree?

0