SISA CSPAI Exam - Topic 2 Question 8 Discussion

Actual exam question for SISA's CSPAI exam

Question #: 8
Topic #: 2

In a scenario where Open-Source LLMs are being used to create a virtual assistant, what would be the most effective way to ensure the assistant is continuously improving its interactions without constant retraining?

ATraining a larger proprietary model to replace the open-source LLM

BShifting the assistant to a completely rule-based system to avoid reliance on user feedback.

CImplementing reinforcement learning from human feedback (RLHF) to refine responses based on user input.

DReducing the amount of feedback integrated to speed up deployment.

Show Suggested Answer

Suggested Answer: C

by William at Nov 20, 2025, 09:47 AM

Limited Time Offer

25%

Off

Get Premium CSPAI Questions as Interactive Web-Based Practice Test or PDF

Contribute your Thoughts:

Submit Cancel

Naomi

2 months ago

C allows for adaptation. It’s the way forward!

upvoted 0 times

...

Beata

2 months ago

D could lead to poor quality. Feedback is essential!

upvoted 0 times

...

Carey

2 months ago

A larger proprietary model? That seems risky and costly.

upvoted 0 times

...

Corazon

2 months ago

Agreed! Continuous learning is key for virtual assistants.

upvoted 0 times

...

Shelton

2 months ago

Reducing feedback? That could backfire big time!

upvoted 0 times

...

Tayna

3 months ago

Totally agree with C, user feedback is key for improvement.

upvoted 0 times

...

Malinda

3 months ago

Wait, shifting to a rule-based system? That sounds limiting!

upvoted 0 times

...

Willodean

3 months ago

A seems like a waste of resources, why replace it?

upvoted 0 times

...

Zena

4 months ago

Haha, D? Really? Let's not half-ass this, folks. C is the way to go!

upvoted 0 times

...

Marica

4 months ago

C is the clear winner here. Who doesn't love a self-improving AI assistant?

upvoted 0 times

...

Cory

4 months ago

Hmm, I'm not sure about C. Seems like a lot of work. Maybe B is the easiest solution?

upvoted 0 times

...

Ty

4 months ago

I agree, C is the best option. Gotta keep that AI assistant learning and improving.

upvoted 0 times

...

Argelia

4 months ago

C is the way to go. Reinforcement learning is the future!

upvoted 0 times

...

Filiberto

4 months ago

I’m not confident, but reducing feedback sounds risky; it could lead to a less effective assistant in the long run.

upvoted 0 times

...

Kristofer

5 months ago

I feel like we practiced a question similar to this, and I think shifting to a rule-based system might limit flexibility.

upvoted 0 times

...

Carlee

5 months ago

Option C definitely seems like the most effective solution. Reinforcement learning allows the assistant to adapt and improve based on real-world interactions, which is crucial for providing a high-quality user experience.

upvoted 0 times

...

Dong

5 months ago

Hmm, I'm not sure about that. Reducing the amount of feedback seems like it would make the assistant less responsive to user needs. I'm leaning more towards option C, but I'll have to think it through a bit more.

upvoted 0 times

...

Malcom

5 months ago

I remember we discussed how RLHF can help models learn from user interactions, but I'm not entirely sure if it's the best choice here.

upvoted 0 times

...

Craig

5 months ago

C is definitely the way to go! RLHF is super effective.

upvoted 0 times

...

Alana

5 months ago

I think C is the best choice. RLHF can really enhance user experience.

upvoted 0 times

...

Phil

6 months ago

I think option C makes sense since it allows for continuous improvement, but I wonder if it requires a lot of resources to implement effectively.

upvoted 0 times

...

Donte

6 months ago

B sounds too rigid. We need flexibility in responses.

upvoted 0 times

...

Herminia

6 months ago

I'm a bit confused on this one. Wouldn't a rule-based system be more reliable than relying on user feedback? I'm not sure if reinforcement learning is the best approach.

upvoted 0 times

...

Lawrence

6 months ago

I think option C is the way to go here. Reinforcement learning from human feedback seems like the most effective way to continuously improve the assistant without constant retraining.

upvoted 0 times