AWS Certified Generative AI Developer - Professional

Get started today

Ultimate access to all questions.

Explanation:

Explanation

The correct answers are B and C because they directly address hallucination reduction while maintaining high throughput and low latency.

Option B (RAG with Amazon Bedrock knowledge base):

Reduces hallucinations at their source by grounding model outputs in verified content through Retrieval Augmented Generation (RAG)
Semantic chunking ensures long technical documents are broken into meaningfully coherent sections
Tuned embeddings allow the model to retrieve only the most relevant chunks rather than processing entire documents in one pass
Significantly improves factual accuracy and reduces cognitive overload on the model
Scales efficiently and supports processing more than 1,000 documents per hour

Option C (Amazon Bedrock guardrails):

Proactively blocks outputs that match known hallucination patterns
Adds an additional layer of safety to ensure generated content adheres to factual constraints
Works alongside RAG to provide comprehensive hallucination reduction

Why other options are incorrect:

Option A (Zero-shot CoT): While chain-of-thought reasoning can improve accuracy, requiring explicit fact verification for each step would significantly increase latency and likely exceed the 3-second requirement
Option D (Increase temperature): Increasing temperature actually increases randomness and creativity, which would likely worsen hallucinations rather than reduce them
Option E (Summarize full document in one pass): This is the company's current problematic approach that causes hallucinations; continuing this method won't solve the problem

Together, options B and C provide a robust, scalable solution that reduces hallucinations while maintaining the required throughput (1,000+ documents/hour) and latency (under 3 seconds per document).

Explanation:

Explanation

The correct answers are B and C because they directly address hallucination reduction while maintaining high throughput and low latency.

Option B (RAG with Amazon Bedrock knowledge base):

Reduces hallucinations at their source by grounding model outputs in verified content through Retrieval Augmented Generation (RAG)
Semantic chunking ensures long technical documents are broken into meaningfully coherent sections
Tuned embeddings allow the model to retrieve only the most relevant chunks rather than processing entire documents in one pass
Significantly improves factual accuracy and reduces cognitive overload on the model
Scales efficiently and supports processing more than 1,000 documents per hour

Option C (Amazon Bedrock guardrails):

Proactively blocks outputs that match known hallucination patterns
Adds an additional layer of safety to ensure generated content adheres to factual constraints
Works alongside RAG to provide comprehensive hallucination reduction

Why other options are incorrect:

Option A (Zero-shot CoT): While chain-of-thought reasoning can improve accuracy, requiring explicit fact verification for each step would significantly increase latency and likely exceed the 3-second requirement
Option D (Increase temperature): Increasing temperature actually increases randomness and creativity, which would likely worsen hallucinations rather than reduce them
Option E (Summarize full document in one pass): This is the company's current problematic approach that causes hallucinations; continuing this method won't solve the problem

Together, options B and C provide a robust, scalable solution that reduces hallucinations while maintaining the required throughput (1,000+ documents/hour) and latency (under 3 seconds per document).

Comments (0)

No comments yet.

A company uses Amazon Bedrock to generate technical content for customers. The company has recently experienced a surge in hallucinated outputs when the company's model generates summaries of long technical documents. The model outputs include inaccurate or fabricated details. The company's current solution uses a large foundation model (FM) with a basic one-shot prompt that includes the full document in a single input.

The company needs a solution that will reduce hallucinations and meet factual accuracy goals. The solution must process more than 1,000 documents each hour and deliver summaries within 3 seconds for each document.

Which combination of solutions will meet these requirements? (Select TWO.)

Real Exam

Community

DDucse

Last updated: March 23, 2026 at 10:35

Implement zero-shot chain-of-thought (CoT) instructions that require step-by-step reasoning with explicit fact verification before the model generates each summary.

0.0%