Total No.
of Questions: 25 x 2 = 50 Marks
Marks eligible to get certification: 70%
Red Colored is Correct answer.
Options may be shuffled in the portal. Read and
understand each answer carefully, then select the correct
one answer.
Don’t by relying on labels like A, B, C, or D.
Here you have only 21 questions with answer, remaining
4 questions answer by your own.
Identifying the function of a tokenizer in Large Language Models involves:
A. Storing linguistic patterns in a structured format
B. Segmenting input text into tokens for model processing
C. Filtering unnecessary data during training
D. Enhancing the computational efficiency of the model
A significant limitation of Generative AI in education involves:
A. The ability to create personalized and adaptive learning environments
B. The generation of biased or contextually inaccurate content
C. The enhancement of accessibility for diverse learners
D. The automation of grading and feedback systems
The application of Generative AI in competency-based learning is best described by:
A. Automating the creation of personalized learning materials based on
individual progress
B. Generating assessments that measure learners' competency levels
C. Enhancing instructor-led training sessions through AI-generated content
D. Facilitating the development of adaptive learning paths for each student
The percentage of GDP targeted for spending on education according to NEP
2020 is:
A. 2%
B. 4%
C. 6%
D. 8%
Arjun, a student, has completed courses at one university but wishes to
transfer his credits to another institution for further studies. The system
introduced in NEP 2020 allows this without losing any progress. What is the
name of this system, and what does it primarily aim to achieve?
A. National Credit Transfer System, enabling seamless credit transfer
between institutions
B. Academic Bank of Credits, facilitating credit transfer across different
institutions
C. University Credit Management System, standardizing educational
curriculums
D. Vocational Credit Transfer System, enhancing vocational training
opportunities
Select the option that most accurately describes how AI tool Synthesia
utilizes deep-learning networks to
A) create natural-sounding avatars that mimic speech and facial
movements
B) automate the video editing process with pre-recorded audio
C) render real-time video content with facial recognition
D) assist in manual avatar animation for video production
A media company creates a video in which a famous actor appears to
endorse a product, but the actor never actually filmed the endorsement.
Instead, the video is generated using artificial intelligence to mimic the
actor's appearance and voice. Which of the following best defines the
technology used in this scenario?
A Face Swap Technology
B Deepfake Technology
C AR VR Technology
D motion capture Technology
Select the option that most accurately explains Retrieval-Augmented
Generation (RAG):
A) A method combining retrieval of external information with the
generative capabilities of an AI model to produce more accurate and
relevant responses
B) A technique where a model generates responses without any external
information, relying solely on pre-trained knowledge
C) A framework enhancing the training of generative models by increasing
the size of training datasets
D) A strategy using retrieval-based models to answer queries without any
generative components
A developer needs to build and run language models on their local
machine that allows easy customization.Which of the following best
describes the tool suitable for this task?
A) Ollama, a lightweight, extensible framework for building and running language
models on the local machine
B) CloudLang – A cloud-based platform for running large-scale language
models with no local machine requirements.
C) DistriTrain – A heavy-duty system for training language models on
distributed networks.
D) LangBridge – A library for integrating language models into third-party
applications without local execution.
Which of the following are practices identified by Google to reduce the
energy and carbon footprints of AI systems?
A) Using sparse models, specialized processors, cloud-based data centers,
and optimizing location
B) Using dense models, general-purpose processors, local data centers,
and global location optimization
C) Relying on high-energy consumption models, general-purpose
processors, and centralized data centres
D) Focusing on increasing model parameters, using non-specialized
processors, and local data centers
Which of the following best describes the two components of the carbon
footprint of a large language model?
A) Operational and Embodied Footprint
B) Software and Data Storage Footprints
C) Data Transfer and Model Training Footprints
D) Cloud-Based and User Device Footprints
Select the option that best highlights the societal impact of generative AI:
A) Automation of jobs, enhanced productivity, and creation of new
economic opportunities
B) The potential to transform industries, leading to new forms of
entertainment and media
C) Enhancement of creativity and support for educational initiatives across
sectors
D) Increased risk of misinformation, bias, and ethical issues in AI-
generated content
Which of the following highlights the use cases of large language models
(LLMs) in the finance sector?
A) Fraud detection and identifying suspicious activities
B) Automated customer support through chatbots
C) Analyzing and extracting insights from financial documents
D) All of the above
Which of the following best explains how generative AI contributes to
sustainability?
A) By optimizing energy consumption, reducing carbon footprints , and
improving supply chain efficiency
B) By increasing energy demands and contributing to higher carbon
emissions
C) By automating unsustainable processes in industries such as
manufacturing and mining
D) By promoting the development of products that solely focus on
entertainment and non-environmental benefits
Select the three popular techniques for implementing Generative AI.
A) Generative Adversarial Networks (GANs), Variational Autoencoders
(VAEs), and Transformers
B) Convolutional Neural Networks (CNNs), Recurrent Neural Networks
(RNNs), and Decision Trees
C) Support Vector Machines (SVMs), K-Means Clustering, and Naive Bayes
D) Gradient Boosting Machines (GBMs), Random Forests, and Principal
Component Analysis (PCA)
Which of the following best describes the role of Transformer architecture
in large language models (LLMs)?
A) It serves as the backbone of LLMs, facilitating efficient handling of long-
range dependencies and parallel processing
B) It is primarily used for image recognition and computer vision tasks
C) It focuses on improving data storage capabilities in AI systems
D) It is designed for sequential data processing without the ability to
parallelize tasks
Which of the following best describes the role of generative AI in
sustainability through smart transportation and e-cities?
A) Optimizing traffic flow, reducing energy consumption, and enhancing
the efficiency of public transportation systems
B) Enabling smart grids and energy-efficient buildings to reduce carbon
emissions in urban areas
C) Generating innovative solutions for waste management and resource
optimization in cities
D) All of the above
Identify the key purpose of Chain-of-Thought (CoT) prompting in natural
language processing.
A. To simplify the language model by reducing its reasoning capabilities
B. To break complex tasks into intermediate reasoning steps for better
understanding
C. To limit the creativity of the model by following fixed rules
D. To increase the model's training data by introducing additional datasets
What is the primary advantage of using Generative AI for simulating
environmental
scenarios?
A. High computational cost
B. Real-time generation of hypothetical data
C. Limited scalability for large datasets
D. Dependency on human intervention
What is a primary application of autoencoders in generative AI related to
image synthesis?
A. Predicting the future positions of objects in video streams
B. Generating realistic images from latent representations
C. Classifying images into predefined categories
D. Detecting anomalies in video frames
Which of the following scenarios demonstrates AI-generated
misinformation?
A. An AI system generating fake social media posts impersonating a
political leader
B. A deep learning model fabricating statistics to falsely support a
controversial argument
C. A chatbot occasionally providing outdated information to user queries
D. An AI-generated image of a non-existent celebrity circulated as real on
social media
Apply your understanding of Variational Autoencoders (VAEs) to determine
their unique functionality.
A Compress data using deterministic techniques to reconstruct the original
input
B. Merge autoencoders with probabilistic models to learn latent data
distributions
C. Focus exclusively on reducing noise in input data through reconstruction
D. Classify data samples into predefined categories using supervised
learning methods
