What is the term for an LLM's ability to continue generating text after receiving a partial input? Please choose the correct answer.

The correct answer is B. Auto-completion. Auto-completion (B) refers to an LLM's core ability to predict and generate the most likely continuation of a given input sequence - when you provide partial text, the model "completes" it based on patterns learned during training. Data Augmentation (A) is a training technique…

Large Language Models (LLMs)

Question

Options

AData Augmentation
BAuto-completion
CZero-shot Learning
DText Extrapolation

How the community answered

(44 responses)

A
9% (4)
B
84% (37)
C
5% (2)
D
2% (1)

Explanation

Auto-completion (B) refers to an LLM's core ability to predict and generate the most likely continuation of a given input sequence - when you provide partial text, the model "completes" it based on patterns learned during training. Data Augmentation (A) is a training technique for expanding datasets, not a generation behavior. Zero-shot Learning (C) describes a model's ability to handle tasks it was never explicitly trained on, unrelated to text continuation. Text Extrapolation (D) sounds plausible but is not a standard NLP term - extrapolation belongs more to numerical/statistical domains.

Memory tip: Think of autocomplete on your phone keyboard - it suggests the next word from what you've already typed. LLMs do the same thing, just much more powerfully. "Auto" = automatic, "completion" = finishing what's started.

Topics

#LLM auto-completion#text generation#partial input continuation

Community Discussion

No community discussion yet for this question.

Full C_AIG_2412 Practice