General MCQs

1372 questions available

FILTER BY SUBTOPIC

All Other 1372
Showing 581–600 of 1372

Q581.Which statement is most accurate about AI-generated deepfake media?

Easy
A It is always harmless
B It can be used for impersonation, fraud, and misinformation
C It is always easy to detect
D It cannot affect investigations
Other PPSC

Correct Answer: B

Deepfakes pose serious security and ethical risks, as they can be used for financial fraud, spreading misinformation, and impersonating individuals.

Q582.Which is the best practice when using AI-generated summaries of seized documents?

Easy
A Accept them without checking
B Compare summaries with original documents to verify details
C Delete original documents to save space
D Use summaries as final evidence directly
Other PPSC

Correct Answer: B

AI summaries can omit or hallucinate critical details; therefore, they must always be verified against the original evidence.

Q583.Which LLM setting is most suitable for consistent, factual answers?

Easy
A High temperature
B Low temperature
C Random temperature
D Maximum creativity mode
Other PPSC

Correct Answer: B

Lowering the temperature setting in an LLM minimizes randomness and creativity, producing more focused and consistent factual responses.

Q584.Which method best reduces hallucination in an LLM used for official work?

Easy
A Higher temperature
B Grounding answers in verified documents or reference materials
C Asking vague questions
D Removing all context
Other PPSC

Correct Answer: B

Grounding responses in verified sources or documents (such as through RAG) ensures that the AI relies on actual facts rather than training weights.

Q585.Which is the safest use of an LLM in case analysis?

Easy
A Replacing the investigator completely
B Assisting with summarization while human investigators verify results
C Creating synthetic evidence
D Ignoring chain of custody
Other PPSC

Correct Answer: B

LLMs are best utilized as tools to accelerate document processing, but human review must remain the final step to guarantee accuracy.

Q586.In an official investigation, why should LLM-generated legal references be verified?

Easy
A LLMs are always offline
B LLMs may hallucinate convincing but fake citations
C LLMs cannot read text
D LLMs only process images
Other PPSC

Correct Answer: B

LLMs are prone to hallucinating citations or legal references that appear highly authentic but are completely fabricated.

Q587.What is the plagiarism risk in using Generative AI?

Easy
A Copying AI-generated work without proper review or attribution
B Cleaning data
C Translating text
D Compressing files
Other PPSC

Correct Answer: A

Plagiarism risk occurs when a user passes off AI-generated output as their own work without original thought or attribution.

Q588.Which practice helps reduce hallucinations in LLM systems?

Easy
A Asking vague questions
B Requesting sources or using retrieved grounding documents
C Increasing randomness
D Ignoring context
Other PPSC

Correct Answer: B

Restricting the model's focus to verified grounding documents prevents it from making up false answers.

Q589.What is a major limitation of LLMs?

Easy
A They can never make mistakes
B They may generate outdated or incorrect information
C They do not process text
D They only work offline
Other PPSC

Correct Answer: B

LLMs can fail on real-time data, complex reasoning, and may present outdated training knowledge as current fact.

Q590.Why is RAG useful?

Easy
A It reduces the need for verification
B It improves factual grounding and reduces hallucinations
C It removes all errors
D It replaces all humans
Other PPSC

Correct Answer: B

By retrieving relevant information from verified documents, RAG ensures that the AI answers are strictly grounded in fact.

Q591.Which of the following is a safe use of Generative AI?

Easy
A Creating fake evidence
B Drafting a study summary
C Spreading misinformation
D Making forged documents
Other PPSC

Correct Answer: B

Using AI to compile study notes, summarize textbooks, or organize ideas are safe and highly productive applications.

Q592.What is a context window in an LLM?

Easy
A Screen size
B The amount of text the model can process and consider at once
C Keyboard layout
D Storage space
Other PPSC

Correct Answer: B

The context window represents the maximum number of tokens a model can read and process in a single prompt and response cycle.

Q593.Which temperature is better for factual exam answers?

Easy
A Very high
B Low
C Random
D Maximum
Other PPSC

Correct Answer: B

A low temperature (e.g., 0.1 or 0) forces the model to choose highly probable, factual, and consistent responses, avoiding creative guesses.

Q594.What does WWW stand for?

Easy
A World Whole Web
B Wide World Web
C Web World Wild
D World Wide Web
Other PPSC

Correct Answer: D

WWW stands for the World Wide Web, an information system created by Tim Berners-Lee in 1989.

Q595.Which of the following acts as temporary storage for content copied or cut?

Easy
A Header
B Sidebar
C Clipboard
D Toolbar
Other PPSC

Correct Answer: C

The clipboard is a temporary operating system buffer that stores copied or cut text, images, or files until they are pasted.

Q596.What is the main goal of Computer Vision?

Easy
A To translate text
B To store files
C To hear sound
D To enable machines to interpret and understand visual data
Other PPSC

Correct Answer: D

Computer Vision is a field of AI that trains computers to interpret and understand the visual world (images and videos).

Q597.The 'GPT' in ChatGPT stands for?

Easy
A Graphical Pattern Tracker
B Global Prediction Technique
C General Processing Tool
D Generative Pre-trained Transformer
Other PPSC

Correct Answer: D

GPT stands for Generative Pre-trained Transformer, an advanced transformer-based neural network architecture optimized for language tasks.

Q598.An expert system mainly consists of a knowledge base and which of the following?

Easy
A Spreadsheet
B Inference engine
C Web server
D Compiler
Other PPSC

Correct Answer: B

The inference engine is the brain of an expert system, applying logical rules to the knowledge base to deduce solutions.

Q599.Which loss function is most commonly used for classification problems?

Easy
A Euclidean distance
B R-squared
C Mean Squared Error
D Cross-Entropy
Other PPSC

Correct Answer: D

Cross-Entropy loss (log loss) measures the performance of a classification model whose output is a probability value between 0 and 1.

Q600.In a GAN, the two main competing components are the generator and the?

Easy
A Optimizer
B Discriminator
C Encoder
D Classifier
Other PPSC

Correct Answer: B

The discriminator acts as a classifier, trying to distinguish between real training data and synthetic data produced by the generator.