AWS Certified AI Practitioner AIF-C01 Practice Question
When preparing data for reinforcement learning from human feedback (RLHF) to fine-tune a text-generation foundation model, which additional dataset is required beyond the data used for ordinary supervised fine-tuning?
An expanded unsupervised corpus of domain-specific documents
A set of labeled demonstration examples showing the correct response for each prompt
A synthetic dataset generated automatically by the model and filtered for quality
Human-ranked pairs of candidate model outputs that indicate relative preference
RLHF first trains the model with standard supervised fine-tuning on high-quality demonstrations, but it then needs a reward model to judge answer quality. The reward model is trained on human preference comparisons-pairs of model outputs that annotators rank better or worse. Supervised fine-tuning does not require these ranked preference pairs; it only uses demonstration input-output examples. Extra unlabeled text or synthetic data can enlarge training corpora but cannot substitute for the preference ranking dataset needed by RLHF.
Ask Bash
Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.
Why are human-ranked preference pairs essential for RLHF?
Open an interactive chat with Bash
What is the role of the reward model in RLHF?
Open an interactive chat with Bash
How does RLHF differ from ordinary supervised fine-tuning?
Open an interactive chat with Bash
AWS Certified AI Practitioner AIF-C01
Applications of Foundation Models
Your Score:
Report Issue
Bash, the Crucial Exams Chat Bot
AI Bot
Loading...
Loading...
Loading...
Pass with Confidence.
IT & Cybersecurity Package
You have hit the limits of our free tier, become a Premium Member today for unlimited access.
Military, Healthcare worker, Gov. employee or Teacher? See if you qualify for a Community Discount.
Monthly
$19.99 $11.99
$11.99/mo
Billed monthly, Cancel any time.
$19.99 after promotion ends
3 Month Pass
$44.99 $26.99
$8.99/mo
One time purchase of $26.99, Does not auto-renew.
$44.99 after promotion ends
Save $18!
MOST POPULAR
Annual Pass
$119.99 $71.99
$5.99/mo
One time purchase of $71.99, Does not auto-renew.
$119.99 after promotion ends
Save $48!
BEST DEAL
Lifetime Pass
$189.99 $113.99
One time purchase, Good for life.
Save $76!
What You Get
All IT & Cybersecurity Package plans include the following perks and exams .