SISA CSPAI Exam - Topic 5 Question 14 Discussion

Actual exam question for SISA's CSPAI exam

Question #: 14
Topic #: 5

Fine-tuning an LLM on a single task involves adjusting model parameters to specialize in a particular domain. What is the primary challenge associated with fine tuning for a single task compared to multi task fine tuning?

ASingle-task fine-tuning introduces more complexity in managing different versions of the model compared to multi-task fine-tuning.

BSingle-task fine-tuning is less effective in generalizing to new, unseen tasks compared to multi-task fine-tuning.

CSingle-task fine-tuning requires significantly more data to achieve comparable performance to multi-task fine tuning.

DSingle-task fine-tuning tends to degrade the model's performance on the original tasks it was trained on.

Show Suggested Answer

Suggested Answer: B

by Noel at May 02, 2026, 12:24 AM

Limited Time Offer

25%

Off

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

Contribute your Thoughts:

Submit Cancel

Kristal

17 days ago

I practiced a question similar to this, and I think single-task fine-tuning does require more data, which might support option C.

upvoted 0 times

...

Lanie

22 days ago

I think I read somewhere that single-task fine-tuning can struggle with generalization, so option B sounds plausible, but I'm not entirely sure.

upvoted 0 times

...

Kallie

27 days ago

I remember discussing how single-task fine-tuning might lead to overfitting, which could relate to option D about degrading performance on original tasks.

upvoted 0 times

...