Fundamentals of Large Language Models →hardMultiple ChoiceObjective-mapped

1Z0-1127-25 Fundamentals of Large Language Models Practice Question

An engineer sets beam search width to 1 during inference on OCI Generative AI. What is the most likely effect on output?

⚠ Common exam trap

A common misconception is that a smaller beam width always degrades quality, but the trap here is that beam width 1 (greedy decoding) is actually the fastest inference method. While it sacrifices diversity and sometimes quality, it is not necessarily worst for all tasks, especially when speed is prioritized.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Faster inference

Beam search width of 1 is equivalent to greedy decoding, where only the single most probable token is selected at each step. This eliminates the need to maintain and compare multiple candidate sequences, significantly reducing computational overhead and memory access, which directly speeds up inference.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✗
More memory usage
Why it's wrong here
Greedy decoding uses less memory as it does not maintain multiple hypotheses.
✗
More diverse outputs
Why it's wrong here
Greedy decoding reduces diversity.
✗
Better quality
Why it's wrong here
Greedy decoding often yields lower quality than beam search with larger width.
✓
Faster inference
Why this is correct
Greedy decoding is the fastest decoding method as it considers only one candidate path.

Go deeper

1Z0-1127-25 Fundamentals of Large Language Models Practice Question

Why each option matters

Option-by-option breakdown

Related to this question