Think Like an AI
Adding more examples from different students would likely improve this.
What is accuracy?
Neural networks were inspired by this organ.
What is the brain?
Machine learning systems learn from these instead of being programmed with rules.
What is data?
ChatGPT is an example of this type of AI.
What is text?
If an AI is trained mostly on one type of data, it may develop this problem.
What is bias?
The first layer of a neural network receives information from the outside world.
What is the input layer?
A collection of examples used for learning.
What is a dataset?
Both Quick Draw and ChatGPT learn from these.
What is data?
When the AI guessed "cat" before you finished drawing, it was recognising this.
What is a pattern?
These layers help the network identify patterns and relationships.
What are hidden layers?
This process improves a model by showing it many examples.
What is training?
The technology that powers ChatGPT is based on very large neural networks called these.
What are Large Language Models?
Traditional programming follows rules written by humans. Machine learning learns patterns from data.
What is the difference between machine learning and traditional programming.
The final prediction is produced by this layer.
What is the output layer?
A model that performs well on new data demonstrates this quality.
What is generalisation?
Unlike Quick Draw, ChatGPT mainly works with this rather than images.
What is language/text?
Quick Draw uses this type of machine learning model.
What is a neural network?
These values determine the strength of connections between neurons.
What are weights?
A model that memorises training examples instead of learning patterns suffers from this problem.
What is overfitting?