Courseiva
hardMultiple ChoiceObjective-mapped

AIF-C01 Practice Question: A media company uses Amazon Bedrock to generate…

A media company uses Amazon Bedrock to generate image captions. They notice that the output quality degrades when the input image contains text in non-Latin scripts. Which model type is MOST likely being used, and what is the likely cause?

⚠ Common exam trap

Many candidates assume all multimodal models handle OCR uniformly, but the exam tests awareness that vision-language models have varying script-specific training biases, and that 'degradation with non-Latin scripts' is a hallmark of OCR weakness, not a generic resolution or modality issue.

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

Anthropic Claude 3 Sonnet; its OCR performance on non-Latin scripts is limited

Anthropic Claude 3 Sonnet is a multimodal model that can process images and generate captions, but its vision encoder relies on OCR-like capabilities that are optimized for Latin scripts. Non-Latin scripts (e.g., Chinese, Arabic, Devanagari) often have complex character shapes and spacing that the model's training data underrepresents, leading to degraded caption quality.

Answer analysis

Option-by-option breakdown

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

  • Anthropic Claude 3 Sonnet; its OCR performance on non-Latin scripts is limited

    Why this is correct

    Claude 3 Sonnet is multimodal and can caption images, but its OCR for non-Latin scripts may be weak, leading to degraded caption quality.

  • Amazon Titan Image Generator; it is not designed for caption generation

    Why it's wrong here

    Titan Image Generator creates images, not captions; it would not be used for this task.

  • Meta Llama 3 70B; it does not accept image inputs

    Why it's wrong here

    Llama 3 is a text-only model and cannot process image inputs for caption generation.

  • Anthropic Claude 3 Sonnet; the image resolution is too low for its vision encoder

    Why it's wrong here

    Resolution is not mentioned as a problem; the issue is specific to non-Latin scripts, pointing to OCR limitations.

About these practice questions

This AIF-C01 question is part of Courseiva's 619-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →

How Courseiva writes practice questions · Editorial policy

JA

Written by Johnson Ajibi, MSc IT Security

Senior Network & Security Engineer · founder of Courseiva

This AIF-C01 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AIF-C01 exam.