For those of you selected in class as the “leave+genAI” group, please follow these instructions:
Step 1: Use the generative AI tool(s) of your choice (e.g. ChatGPT, Claude, Gemini) for roughly 15 minutes to teach yourself about the following topics relating to neural network overfitting:¶
regularization
dropout
data augmentation
Interact with the genAI tools however you’d like during that 15 minutes. Your goal: maximally understand the overall conceptual basis for each of these techniques. Do not worry about mathematical formalism, specific technical implementations, etc. Focus on the big picture.
Only spend 15 minutes, because that’s the amount of time the other students, who stayed in class, will spend on it as well.
Step 2: Immediately after the 15 minutes is up, complete the quiz at this Google form. Take no more than 10 minutes on it.¶
IMPORTANT: use genAI for the ~15 minute learning session, but NOT for the quiz. Use no external resources on the quiz in fact: pretend it was just like our midterm.
You will not be penalized for wrong answers, but please still try to do your very best