Free NVIDIA NCA-GENL Actual Exam Questions - Question 3 Discussion

Question No. 3
Why do we need positional encoding in transformer-based models?
Select one option, then reveal solution.
US
CE
Carlos E.
2026-02-19

Maybe D doesn’t make much sense since positional encoding doesn’t speed up processing. It’s definitely not about reducing dimensionality (C) or stopping overfitting (B), so A still feels right.

0
CE
Carlos E.
2026-02-18

Probably A, since transformers lack any built-in sense of sequence order.

0
SR
Sam R.
2026-01-31

A, transformers don’t inherently understand sequence order without it.

0
NP
Naveed P.
2026-01-29

It’s A because transformers lack recurrence or convolution, so they need position info.

0
NP
Naveed P.
2026-01-27

Makes sense it’s about order since transformers treat inputs as sets, so A.

0
PM
Paul M.
2026-01-24

Positional encoding is definitely about order, since transformers don’t have any built-in notion of sequence position. So it’s not about reducing dimensions or speeding up anything, ruling out C and D.

0
SH
Saad H.
2026-01-16

Makes sense that order matters here, so A.

0