NEW
Font size
WorksheetsAgents Unleashed: Quiz on LLMs & AI Frontiers
Total questions: 30
Worksheet time: 15mins
What does the acronym LLM stand for in the context of AI?
Linear Logic Model
Large Language Model
Limited Learning Machine
Language Learning Mechanism
Which of the following is NOT a current trend in the development of LLMs?
Multimodal capabilities
On-device deployment
Quantum computing integration
Code generation
What is the main goal of an agentic system in AI?
To translate languages
To perform simple tasks on command
To act autonomously towards a goal
To compress data efficiently
Which of the following is a popular open-source LLM framework?
TensorBoard
LangChain
GitHub Copilot
Visual Studio
In the context of agentic AI systems, what is a “tool use” capability?
The ability to install software
The ability to interact with external APIs or environments
The ability to design neural networks
The ability to repair software bugs
Which recent LLM from OpenAI supports multimodal inputs (text, image, audio)?
GPT-3
ChatGPT
GPT-4-turbo
GPT-Neo
What is a significant challenge with deploying agentic systems?
Too much memory usage
Difficulty in writing Java code
Safety, alignment, and hallucination risks
Lack of documentation
Which of the following best describes "chain-of-thought" prompting in LLMs?
Executing multiple threads in parallel
Encouraging the model to list a series of steps or reasoning
Jumping directly to the answer
Forcing the model to restart its answer
Which architecture underlies most modern LLMs, including GPT and BERT?
Recurrent Neural Network
Convolutional Neural Network
Transformer
GAN (Generative Adversarial Network)
What is an "agent loop" in agentic AI systems?
A loop for debugging AI
A repetitive training cycle
A cycle of observe → reason → act
A recursive error handling mechanism
What is one key trend in the evolution of LLMs?
Smaller datasets
Limited access
On-device fine-tuning
Removal of AI safety layers
Which of these tasks can modern LLMs perform?
Language translation
Code generation
Document summarization
All of the above
What does "multimodal" mean in LLMs?
Supporting multiple users
Understanding images, text, and audio
Running on multiple devices
Generating multiple outputs
Which LLM was announced with a "memory" feature to retain chat history over time?
Claude
ChatGPT
Gemini
BERT
What is fine-tuning in the context of LLMs?
Adding noise to models
Training on new data to specialize
Compressing the model
Testing model speed
What defines an agentic AI system?
Pre-trained for customer support
Rule-based chatbot
Autonomously takes actions toward a goal
Only responds to prompts
Which of these is an example of an LLM agent toolchain?
PyTorch
LangChain
TensorBoard
VS Code
What is a “tool use” capability in agentic systems?
Writing code
Searching the web or invoking APIs
Using keyboard shortcuts
Switching UI themes
Which prompt technique helps LLMs reason step-by-step?
Zero-shot
Few-shot
Chain-of-Thought
Rapid-Fire
Which of the following is a key concern in agentic AI systems?
Slow response time
A circular network structureMemory leaks
Safety and hallucinations
Low internet speed
What is a common method to ensure safety in agentic AI systems?
Conducting thorough testing and validation
Implementing strict access controls
Regular software updates
Using outdated algorithms
Which of the following describes a potential benefit of agentic systems?
Reduced computational power requirements
Enhanced automation of complex tasks
Limited functionality
Increased human oversight
What is a primary focus when developing agentic AI systems?
Minimizing user interaction
Maximizing data storage
Ensuring ethical decision-making
Improving graphical user interfaces
What is a common risk associated with the use of agentic AI systems?
Enhanced user engagement
Data privacy concerns
Increased operational costs
Faster processing speeds
Which of the following is a method to improve the performance of LLMs?
Using outdated algorithms
Limiting input types
Increasing training data diversity
Reducing model size
What is the purpose of reinforcement learning in the context of agentic systems?
To enhance user interface design
To optimize decision-making through feedback
To reduce computational load
To simplify code writing
What is a potential ethical concern when deploying agentic AI systems?
Lower operational costs
Bias in decision-making
Increased efficiency
Improved user satisfaction
How do agentic AI systems typically learn from their environment?
Through supervised learning only
Through unsupervised clustering
By using reinforcement learning techniques
By manual programming
What is a key feature of agentic AI systems that distinguishes them from traditional AI?
Ability to process large datasets
Capability to make autonomous decisions
Reliance on human input for every action
Focus on data analysis
What is alignment in AI?
What is alignment in AI?
Model's behavior aligning with human intent and ethics
Training a smaller model
Data cleanup
