Enter Title

AI Systems

100

This language model by OpenAI was used as a predecessor to GPT-4.?

GPT-3

200

What is the LLM developed by Google?

Genemi

300

What is the open source large language model developed by Meta?

LLaMA (Large Language Model Meta AI).

400

Which LLM sent U.S Tech stocks crashing this week?

DeepSeek

500

This model architecture introduced by OpenAI excels at generating coherent and context-aware text

GPT

AI Terminology

100

This term refers to AI that can outperform humans in almost every field.

artificial general intelligence (AGI)

200

This term refers to the ability of a machine to mimic human intelligence

artificial intelligence

300

This AI field focuses on teaching machines to interpret and understand visual data

Computer Vision

400

What is a field of artificial intelligence (AI) focused on enabling computers to understand, interpret, generate, and respond to human language in a way that is both meaningful and useful. 

Natural Language Processing

500

This subset of AI is inspired by the structure and function of the human brain.

neural networks

AI Math

100

AI was invented in North Carolina

False

but Airplanes were

200

True or False Calculus is used to find the specific weights to minimize value of the loss function in machine learning?

True

300

In a clustering algorithm, you are given two points:

• Point 1: (2,3)
• Point 2: (5,7)

Calculate the distance between the two points

5

400

A perceptron receives the following inputs and weights:

• Inputs: [1,2,3]
• Weights: [0.5,0.25,0.75]

Calculate the weighted sum (z) using the formula:

Z = ∑ (input * weight)

3.25

500

If a neural network has 3 layers: an input layer with 5 neurons, a hidden layer with 10 neurons, and an output layer with 2 neurons, how many total weights (connections between nuerons) are there if each neuron in one layer is connected to every neuron in the next layer?

70

Connections between input and hidden layer (5*10=50)

Connections between hidden and output layer (10*2=20)

Total Connections = 50+70

AI logic

100

Consider the following conditions for a machine learning model to classify an instance as "Spam":

• Condition 1: The email contains the word "free".
• Condition 2: The email is from a known sender.

The model works as follows:

• If Condition 1 AND Condition 2 are true, classify as "Spam".
• If Condition 1 OR Condition 2 is true (but not both), classify as "Not Spam".

If an email contains the word "free" but is from an unknown sender, how would the model classify it?

Given that the email contains the word "free" (Condition 1) but is from an unknown sender (Condition 2 is false), the model would follow the second rule:

• If Condition 1 OR Condition 2 is true (but not both), the email would be classified as "Not Spam".

So, the email would be classified as "Not Spam".

200

Consider the rule in a smart home system:

• Rule: "If the temperature is above 25°C, turn on the air conditioning."

Which of the following logically follows from this rule?

1. The temperature is above 25°C. Therefore, the air conditioning must be on.
2. The air conditioning is on. Therefore, the temperature must be above 25°C.

1. The temperature is above 25°C. Therefore, the air conditioning must be on.

   • This logically follows from the rule. If the temperature is above 25°C, the rule specifies that the air conditioning should be turned on. So, this is a valid conclusion.

2. The air conditioning is on. Therefore, the temperature must be above 25°C.

   • This does not logically follow from the rule. The rule only specifies what happens when the temperature is above 25°C, but it doesn't state that the air conditioning can only be on when the temperature is above 25°C. There could be other reasons for the air conditioning to be on (e.g., a manual override or another condition), so this is not a valid conclusion.

So, the first statement is the one that logically follows from the rule.

300

In a reinforcement learning scenario, an agent is in an environment where it needs to maximize its rewards. The agent can take two possible actions:

• Action 1: Gain 10 points but has a 50% chance of losing 5 points.
• Action 2: Gain 5 points but has no risk of losing any points.

Which action should the agent take based on the expected value?

Action 1:

• Gain 10 points with a 50% probability (0.5).
• Lose 5 points with a 50% probability (0.5).

The expected value for Action 1 is calculated as:

Expected Value of Action 1=(.5*10)+(.5*-5)=5-2.5=2.5

Action 2:

• Gain 5 points with 100% probability (1.0).

The expected value for Action 2 is:

Expected Value of Action 2=5

Conclusion:

• Action 1 has an expected value of 2.5 points.
• Action 2 has an expected value of 5 points.

Based on the expected values, the agent should take Action 2, as it provides a higher expected reward (5 points) compared to Action 1 (2.5 points).